Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
// Copyright © 2013-14 Steve Francia <spf@spf13.com>.
|
2013-07-04 11:32:55 -04:00
|
|
|
|
//
|
2015-11-23 22:16:36 -05:00
|
|
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
2013-07-04 11:32:55 -04:00
|
|
|
|
// you may not use this file except in compliance with the License.
|
|
|
|
|
// You may obtain a copy of the License at
|
2015-11-23 22:16:36 -05:00
|
|
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
2013-07-04 11:32:55 -04:00
|
|
|
|
//
|
|
|
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
|
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
|
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
|
|
|
// See the License for the specific language governing permissions and
|
|
|
|
|
// limitations under the License.
|
|
|
|
|
|
|
|
|
|
package hugolib
|
|
|
|
|
|
|
|
|
|
import (
|
2015-06-21 07:08:30 -04:00
|
|
|
|
"bytes"
|
|
|
|
|
"errors"
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
"fmt"
|
2014-01-29 17:50:31 -05:00
|
|
|
|
"html/template"
|
2014-02-25 23:57:31 -05:00
|
|
|
|
"reflect"
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
"regexp"
|
2014-11-18 04:20:52 -05:00
|
|
|
|
"sort"
|
2014-01-29 17:50:31 -05:00
|
|
|
|
"strings"
|
2015-01-21 13:13:11 -05:00
|
|
|
|
"sync"
|
2014-11-20 12:32:21 -05:00
|
|
|
|
|
2015-01-30 14:19:46 -05:00
|
|
|
|
bp "github.com/spf13/hugo/bufferpool"
|
2014-11-20 12:32:21 -05:00
|
|
|
|
"github.com/spf13/hugo/helpers"
|
|
|
|
|
"github.com/spf13/hugo/tpl"
|
|
|
|
|
jww "github.com/spf13/jwalterweatherman"
|
2014-03-31 13:23:34 -04:00
|
|
|
|
)
|
2013-07-04 11:32:55 -04:00
|
|
|
|
|
|
|
|
|
type ShortcodeWithPage struct {
|
2015-11-20 19:59:54 -05:00
|
|
|
|
Params interface{}
|
|
|
|
|
Inner template.HTML
|
|
|
|
|
Page *Page
|
|
|
|
|
IsNamedParams bool
|
2013-07-04 11:32:55 -04:00
|
|
|
|
}
|
|
|
|
|
|
Provide (relative) reference funcs & shortcodes.
- `.Ref` and `.RelRef` take a reference (the logical filename for a
page, including extension and/or a document fragment ID) and return
a permalink (or relative permalink) to the referenced document.
- If the reference is a page name (such as `about.md`), the page
will be discovered and the permalink will be returned: `/about/`
- If the reference is a page name with a fragment (such as
`about.md#who`), the page will be discovered and used to add the
`page.UniqueID()` to the resulting fragment and permalink:
`/about/#who:deadbeef`.
- If the reference is a fragment and `.*Ref` has been called from
a `Node` or `SiteInfo`, it will be returned as is: `#who`.
- If the reference is a fragment and `.*Ref` has been called from
a `Page`, it will be returned with the page’s unique ID:
`#who:deadbeef`.
- `.*Ref` can be called from either `Node`, `SiteInfo` (e.g.,
`Node.Site`), `Page` objects, or `ShortcodeWithPage` objects in
templates.
- `.*Ref` cannot be used in content, so two shortcodes have been
created to provide the functionality to content: `ref` and `relref`.
These are intended to be used within markup, like `[Who]({{% ref
about.md#who %}})` or `<a href="{{% ref about.md#who %}}">Who</a>`.
- There are also `ref` and `relref` template functions (used to create
the shortcodes) that expect a `Page` or `Node` object and the
reference string (e.g., `{{ relref . "about.md" }}` or `{{
"about.md" | ref . }}`). It actually looks for `.*Ref` as defined on
`Node` or `Page` objects.
- Shortcode handling had to use a *differently unique* wrapper in
`createShortcodePlaceholder` because of the way that the `ref` and
`relref` are intended to be used in content.
2014-11-24 01:15:34 -05:00
|
|
|
|
func (scp *ShortcodeWithPage) Ref(ref string) (string, error) {
|
|
|
|
|
return scp.Page.Ref(ref)
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
func (scp *ShortcodeWithPage) RelRef(ref string) (string, error) {
|
|
|
|
|
return scp.Page.RelRef(ref)
|
|
|
|
|
}
|
|
|
|
|
|
2015-04-19 08:50:24 -04:00
|
|
|
|
func (scp *ShortcodeWithPage) Scratch() *Scratch {
|
|
|
|
|
return scp.Page.Scratch()
|
|
|
|
|
}
|
|
|
|
|
|
2014-02-25 23:57:31 -05:00
|
|
|
|
func (scp *ShortcodeWithPage) Get(key interface{}) interface{} {
|
|
|
|
|
if reflect.ValueOf(scp.Params).Len() == 0 {
|
|
|
|
|
return nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
var x reflect.Value
|
|
|
|
|
|
|
|
|
|
switch key.(type) {
|
|
|
|
|
case int64, int32, int16, int8, int:
|
|
|
|
|
if reflect.TypeOf(scp.Params).Kind() == reflect.Map {
|
|
|
|
|
return "error: cannot access named params by position"
|
|
|
|
|
} else if reflect.TypeOf(scp.Params).Kind() == reflect.Slice {
|
2015-08-07 13:21:26 -04:00
|
|
|
|
idx := int(reflect.ValueOf(key).Int())
|
|
|
|
|
ln := reflect.ValueOf(scp.Params).Len()
|
|
|
|
|
if idx > ln-1 {
|
|
|
|
|
helpers.DistinctErrorLog.Printf("No shortcode param at .Get %d in page %s, have params: %v", idx, scp.Page.FullFilePath(), scp.Params)
|
|
|
|
|
return fmt.Sprintf("error: index out of range for positional param at position %d", idx)
|
|
|
|
|
}
|
|
|
|
|
x = reflect.ValueOf(scp.Params).Index(idx)
|
2014-02-25 23:57:31 -05:00
|
|
|
|
}
|
|
|
|
|
case string:
|
|
|
|
|
if reflect.TypeOf(scp.Params).Kind() == reflect.Map {
|
|
|
|
|
x = reflect.ValueOf(scp.Params).MapIndex(reflect.ValueOf(key))
|
|
|
|
|
if !x.IsValid() {
|
|
|
|
|
return ""
|
|
|
|
|
}
|
|
|
|
|
} else if reflect.TypeOf(scp.Params).Kind() == reflect.Slice {
|
|
|
|
|
if reflect.ValueOf(scp.Params).Len() == 1 && reflect.ValueOf(scp.Params).Index(0).String() == "" {
|
|
|
|
|
return nil
|
|
|
|
|
}
|
|
|
|
|
return "error: cannot access positional params by string name"
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
switch x.Kind() {
|
|
|
|
|
case reflect.String:
|
|
|
|
|
return x.String()
|
|
|
|
|
case reflect.Int64, reflect.Int32, reflect.Int16, reflect.Int8, reflect.Int:
|
|
|
|
|
return x.Int()
|
|
|
|
|
default:
|
|
|
|
|
return x
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
// Note - this value must not contain any markup syntax
|
|
|
|
|
const shortcodePlaceholderPrefix = "HUGOSHORTCODE"
|
2013-07-04 11:32:55 -04:00
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
type shortcode struct {
|
|
|
|
|
name string
|
|
|
|
|
inner []interface{} // string or nested shortcode
|
|
|
|
|
params interface{} // map or array
|
|
|
|
|
err error
|
|
|
|
|
doMarkup bool
|
2013-07-04 11:32:55 -04:00
|
|
|
|
}
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
func (sc shortcode) String() string {
|
|
|
|
|
// for testing (mostly), so any change here will break tests!
|
2014-11-18 04:20:52 -05:00
|
|
|
|
var params interface{}
|
|
|
|
|
switch v := sc.params.(type) {
|
|
|
|
|
case map[string]string:
|
|
|
|
|
// sort the keys so test assertions won't fail
|
|
|
|
|
var keys []string
|
|
|
|
|
for k := range v {
|
|
|
|
|
keys = append(keys, k)
|
|
|
|
|
}
|
|
|
|
|
sort.Strings(keys)
|
|
|
|
|
var tmp = make([]string, len(keys))
|
|
|
|
|
|
|
|
|
|
for i, k := range keys {
|
|
|
|
|
tmp[i] = k + ":" + v[k]
|
|
|
|
|
}
|
|
|
|
|
params = tmp
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
// use it as is
|
|
|
|
|
params = sc.params
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return fmt.Sprintf("%s(%q, %t){%s}", sc.name, params, sc.doMarkup, sc.inner)
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
2013-12-06 23:14:54 -05:00
|
|
|
|
|
2015-04-29 13:08:34 -04:00
|
|
|
|
// HandleShortcodes does all in one go: extract, render and replace
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
// only used for testing
|
2015-04-29 13:08:34 -04:00
|
|
|
|
func HandleShortcodes(stringToParse string, page *Page, t tpl.Template) (string, error) {
|
2015-04-30 09:59:14 -04:00
|
|
|
|
tmpContent, tmpShortcodes, err := extractAndRenderShortcodes(stringToParse, page, t)
|
|
|
|
|
|
|
|
|
|
if err != nil {
|
|
|
|
|
return "", err
|
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
|
|
|
|
if len(tmpShortcodes) > 0 {
|
2015-06-21 07:08:30 -04:00
|
|
|
|
tmpContentWithTokensReplaced, err := replaceShortcodeTokens([]byte(tmpContent), shortcodePlaceholderPrefix, tmpShortcodes)
|
2013-12-06 23:14:54 -05:00
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
if err != nil {
|
2015-04-29 13:08:34 -04:00
|
|
|
|
return "", fmt.Errorf("Fail to replace short code tokens in %s:\n%s", page.BaseFileName(), err.Error())
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
} else {
|
2015-04-29 13:08:34 -04:00
|
|
|
|
return string(tmpContentWithTokensReplaced), nil
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2015-04-29 13:08:34 -04:00
|
|
|
|
return string(tmpContent), nil
|
2013-12-06 23:14:54 -05:00
|
|
|
|
}
|
|
|
|
|
|
2015-01-21 13:13:11 -05:00
|
|
|
|
var isInnerShortcodeCache = struct {
|
|
|
|
|
sync.RWMutex
|
|
|
|
|
m map[string]bool
|
|
|
|
|
}{m: make(map[string]bool)}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
|
|
|
|
// to avoid potential costly look-aheads for closing tags we look inside the template itself
|
|
|
|
|
// we could change the syntax to self-closing tags, but that would make users cry
|
|
|
|
|
// the value found is cached
|
2015-11-21 17:29:58 -05:00
|
|
|
|
func isInnerShortcode(t *template.Template) (bool, error) {
|
2015-01-21 13:13:11 -05:00
|
|
|
|
isInnerShortcodeCache.RLock()
|
|
|
|
|
m, ok := isInnerShortcodeCache.m[t.Name()]
|
|
|
|
|
isInnerShortcodeCache.RUnlock()
|
|
|
|
|
|
|
|
|
|
if ok {
|
2015-11-21 17:29:58 -05:00
|
|
|
|
return m, nil
|
2014-06-26 16:47:51 -04:00
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
2015-01-21 13:13:11 -05:00
|
|
|
|
isInnerShortcodeCache.Lock()
|
2015-02-20 11:41:37 -05:00
|
|
|
|
defer isInnerShortcodeCache.Unlock()
|
2015-11-21 17:29:58 -05:00
|
|
|
|
if t.Tree == nil {
|
|
|
|
|
return false, errors.New("Template failed to compile")
|
|
|
|
|
}
|
2015-02-20 11:41:37 -05:00
|
|
|
|
match, _ := regexp.MatchString("{{.*?\\.Inner.*?}}", t.Tree.Root.String())
|
2015-01-21 13:13:11 -05:00
|
|
|
|
isInnerShortcodeCache.m[t.Name()] = match
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
2015-11-21 17:29:58 -05:00
|
|
|
|
return match, nil
|
2013-12-06 23:14:54 -05:00
|
|
|
|
}
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
func createShortcodePlaceholder(id int) string {
|
Provide (relative) reference funcs & shortcodes.
- `.Ref` and `.RelRef` take a reference (the logical filename for a
page, including extension and/or a document fragment ID) and return
a permalink (or relative permalink) to the referenced document.
- If the reference is a page name (such as `about.md`), the page
will be discovered and the permalink will be returned: `/about/`
- If the reference is a page name with a fragment (such as
`about.md#who`), the page will be discovered and used to add the
`page.UniqueID()` to the resulting fragment and permalink:
`/about/#who:deadbeef`.
- If the reference is a fragment and `.*Ref` has been called from
a `Node` or `SiteInfo`, it will be returned as is: `#who`.
- If the reference is a fragment and `.*Ref` has been called from
a `Page`, it will be returned with the page’s unique ID:
`#who:deadbeef`.
- `.*Ref` can be called from either `Node`, `SiteInfo` (e.g.,
`Node.Site`), `Page` objects, or `ShortcodeWithPage` objects in
templates.
- `.*Ref` cannot be used in content, so two shortcodes have been
created to provide the functionality to content: `ref` and `relref`.
These are intended to be used within markup, like `[Who]({{% ref
about.md#who %}})` or `<a href="{{% ref about.md#who %}}">Who</a>`.
- There are also `ref` and `relref` template functions (used to create
the shortcodes) that expect a `Page` or `Node` object and the
reference string (e.g., `{{ relref . "about.md" }}` or `{{
"about.md" | ref . }}`). It actually looks for `.*Ref` as defined on
`Node` or `Page` objects.
- Shortcode handling had to use a *differently unique* wrapper in
`createShortcodePlaceholder` because of the way that the `ref` and
`relref` are intended to be used in content.
2014-11-24 01:15:34 -05:00
|
|
|
|
return fmt.Sprintf("{@{@%s-%d@}@}", shortcodePlaceholderPrefix, id)
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
|
|
|
|
|
Provide (relative) reference funcs & shortcodes.
- `.Ref` and `.RelRef` take a reference (the logical filename for a
page, including extension and/or a document fragment ID) and return
a permalink (or relative permalink) to the referenced document.
- If the reference is a page name (such as `about.md`), the page
will be discovered and the permalink will be returned: `/about/`
- If the reference is a page name with a fragment (such as
`about.md#who`), the page will be discovered and used to add the
`page.UniqueID()` to the resulting fragment and permalink:
`/about/#who:deadbeef`.
- If the reference is a fragment and `.*Ref` has been called from
a `Node` or `SiteInfo`, it will be returned as is: `#who`.
- If the reference is a fragment and `.*Ref` has been called from
a `Page`, it will be returned with the page’s unique ID:
`#who:deadbeef`.
- `.*Ref` can be called from either `Node`, `SiteInfo` (e.g.,
`Node.Site`), `Page` objects, or `ShortcodeWithPage` objects in
templates.
- `.*Ref` cannot be used in content, so two shortcodes have been
created to provide the functionality to content: `ref` and `relref`.
These are intended to be used within markup, like `[Who]({{% ref
about.md#who %}})` or `<a href="{{% ref about.md#who %}}">Who</a>`.
- There are also `ref` and `relref` template functions (used to create
the shortcodes) that expect a `Page` or `Node` object and the
reference string (e.g., `{{ relref . "about.md" }}` or `{{
"about.md" | ref . }}`). It actually looks for `.*Ref` as defined on
`Node` or `Page` objects.
- Shortcode handling had to use a *differently unique* wrapper in
`createShortcodePlaceholder` because of the way that the `ref` and
`relref` are intended to be used in content.
2014-11-24 01:15:34 -05:00
|
|
|
|
const innerNewlineRegexp = "\n"
|
|
|
|
|
const innerCleanupRegexp = `\A<p>(.*)</p>\n\z`
|
|
|
|
|
const innerCleanupExpand = "$1"
|
|
|
|
|
|
2015-01-15 06:21:15 -05:00
|
|
|
|
func renderShortcode(sc shortcode, p *Page, t tpl.Template) string {
|
2015-04-17 18:40:54 -04:00
|
|
|
|
tmpl := getShortcodeTemplate(sc.name, t)
|
2014-02-25 23:57:31 -05:00
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
if tmpl == nil {
|
|
|
|
|
jww.ERROR.Printf("Unable to locate template for shortcode '%s' in page %s", sc.name, p.BaseFileName())
|
|
|
|
|
return ""
|
|
|
|
|
}
|
2014-02-25 23:57:31 -05:00
|
|
|
|
|
2015-11-20 19:59:54 -05:00
|
|
|
|
data := &ShortcodeWithPage{Params: sc.params, Page: p}
|
|
|
|
|
if sc.params != nil {
|
|
|
|
|
data.IsNamedParams = reflect.TypeOf(sc.params).Kind() == reflect.Map
|
|
|
|
|
}
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
if len(sc.inner) > 0 {
|
|
|
|
|
var inner string
|
|
|
|
|
for _, innerData := range sc.inner {
|
|
|
|
|
switch innerData.(type) {
|
|
|
|
|
case string:
|
|
|
|
|
inner += innerData.(string)
|
|
|
|
|
case shortcode:
|
2015-01-15 06:21:15 -05:00
|
|
|
|
inner += renderShortcode(innerData.(shortcode), p, t)
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
default:
|
|
|
|
|
jww.ERROR.Printf("Illegal state on shortcode rendering of '%s' in page %s. Illegal type in inner data: %s ",
|
|
|
|
|
sc.name, p.BaseFileName(), reflect.TypeOf(innerData))
|
|
|
|
|
return ""
|
2014-02-25 23:57:31 -05:00
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if sc.doMarkup {
|
2015-03-06 08:56:44 -05:00
|
|
|
|
newInner := helpers.RenderBytes(&helpers.RenderingContext{
|
2014-11-28 15:16:57 -05:00
|
|
|
|
Content: []byte(inner), PageFmt: p.guessMarkupType(),
|
2015-03-11 13:34:57 -04:00
|
|
|
|
DocumentID: p.UniqueID(), Config: p.getRenderingConfig()})
|
Provide (relative) reference funcs & shortcodes.
- `.Ref` and `.RelRef` take a reference (the logical filename for a
page, including extension and/or a document fragment ID) and return
a permalink (or relative permalink) to the referenced document.
- If the reference is a page name (such as `about.md`), the page
will be discovered and the permalink will be returned: `/about/`
- If the reference is a page name with a fragment (such as
`about.md#who`), the page will be discovered and used to add the
`page.UniqueID()` to the resulting fragment and permalink:
`/about/#who:deadbeef`.
- If the reference is a fragment and `.*Ref` has been called from
a `Node` or `SiteInfo`, it will be returned as is: `#who`.
- If the reference is a fragment and `.*Ref` has been called from
a `Page`, it will be returned with the page’s unique ID:
`#who:deadbeef`.
- `.*Ref` can be called from either `Node`, `SiteInfo` (e.g.,
`Node.Site`), `Page` objects, or `ShortcodeWithPage` objects in
templates.
- `.*Ref` cannot be used in content, so two shortcodes have been
created to provide the functionality to content: `ref` and `relref`.
These are intended to be used within markup, like `[Who]({{% ref
about.md#who %}})` or `<a href="{{% ref about.md#who %}}">Who</a>`.
- There are also `ref` and `relref` template functions (used to create
the shortcodes) that expect a `Page` or `Node` object and the
reference string (e.g., `{{ relref . "about.md" }}` or `{{
"about.md" | ref . }}`). It actually looks for `.*Ref` as defined on
`Node` or `Page` objects.
- Shortcode handling had to use a *differently unique* wrapper in
`createShortcodePlaceholder` because of the way that the `ref` and
`relref` are intended to be used in content.
2014-11-24 01:15:34 -05:00
|
|
|
|
|
|
|
|
|
// If the type is “unknown” or “markdown”, we assume the markdown
|
|
|
|
|
// generation has been performed. Given the input: `a line`, markdown
|
|
|
|
|
// specifies the HTML `<p>a line</p>\n`. When dealing with documents as a
|
|
|
|
|
// whole, this is OK. When dealing with an `{{ .Inner }}` block in Hugo,
|
|
|
|
|
// this is not so good. This code does two things:
|
|
|
|
|
//
|
|
|
|
|
// 1. Check to see if inner has a newline in it. If so, the Inner data is
|
|
|
|
|
// unchanged.
|
|
|
|
|
// 2 If inner does not have a newline, strip the wrapping <p> block and
|
|
|
|
|
// the newline. This was previously tricked out by wrapping shortcode
|
|
|
|
|
// substitutions in <div>HUGOSHORTCODE-1</div> which prevents the
|
|
|
|
|
// generation, but means that you can’t use shortcodes inside of
|
|
|
|
|
// markdown structures itself (e.g., `[foo]({{% ref foo.md %}})`).
|
|
|
|
|
switch p.guessMarkupType() {
|
|
|
|
|
case "unknown", "markdown":
|
|
|
|
|
if match, _ := regexp.MatchString(innerNewlineRegexp, inner); !match {
|
|
|
|
|
cleaner, err := regexp.Compile(innerCleanupRegexp)
|
|
|
|
|
|
|
|
|
|
if err == nil {
|
|
|
|
|
newInner = cleaner.ReplaceAll(newInner, []byte(innerCleanupExpand))
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
data.Inner = template.HTML(newInner)
|
2014-02-25 23:57:31 -05:00
|
|
|
|
} else {
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
data.Inner = template.HTML(inner)
|
2014-02-25 23:57:31 -05:00
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
2014-02-25 23:57:31 -05:00
|
|
|
|
}
|
|
|
|
|
|
2015-04-17 18:40:54 -04:00
|
|
|
|
return renderShortcodeWithPage(tmpl, data)
|
2014-02-25 23:57:31 -05:00
|
|
|
|
}
|
|
|
|
|
|
2015-04-30 09:59:14 -04:00
|
|
|
|
func extractAndRenderShortcodes(stringToParse string, p *Page, t tpl.Template) (string, map[string]string, error) {
|
2014-01-29 17:50:31 -05:00
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
content, shortcodes, err := extractShortcodes(stringToParse, p, t)
|
|
|
|
|
renderedShortcodes := make(map[string]string)
|
|
|
|
|
|
|
|
|
|
if err != nil {
|
|
|
|
|
// try to render what we have whilst logging the error
|
|
|
|
|
jww.ERROR.Println(err.Error())
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
for key, sc := range shortcodes {
|
|
|
|
|
if sc.err != nil {
|
|
|
|
|
// need to have something to replace with
|
|
|
|
|
renderedShortcodes[key] = ""
|
|
|
|
|
} else {
|
2015-01-15 06:21:15 -05:00
|
|
|
|
renderedShortcodes[key] = renderShortcode(sc, p, t)
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
|
|
|
|
}
|
2014-01-29 17:50:31 -05:00
|
|
|
|
|
2015-04-30 09:59:14 -04:00
|
|
|
|
return content, renderedShortcodes, err
|
2014-02-25 23:57:31 -05:00
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
|
|
|
|
|
2015-08-07 14:08:23 -04:00
|
|
|
|
var shortCodeIllegalState = errors.New("Illegal shortcode state")
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
// pageTokens state:
|
|
|
|
|
// - before: positioned just before the shortcode start
|
|
|
|
|
// - after: shortcode(s) consumed (plural when they are nested)
|
2014-11-20 12:32:21 -05:00
|
|
|
|
func extractShortcode(pt *pageTokens, p *Page, t tpl.Template) (shortcode, error) {
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
sc := shortcode{}
|
|
|
|
|
var isInner = false
|
|
|
|
|
|
|
|
|
|
var currItem item
|
|
|
|
|
var cnt = 0
|
|
|
|
|
|
|
|
|
|
Loop:
|
|
|
|
|
for {
|
|
|
|
|
currItem = pt.next()
|
|
|
|
|
|
|
|
|
|
switch currItem.typ {
|
|
|
|
|
case tLeftDelimScWithMarkup, tLeftDelimScNoMarkup:
|
|
|
|
|
next := pt.peek()
|
|
|
|
|
if next.typ == tScClose {
|
|
|
|
|
continue
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
if cnt > 0 {
|
|
|
|
|
// nested shortcode; append it to inner content
|
|
|
|
|
pt.backup3(currItem, next)
|
|
|
|
|
nested, err := extractShortcode(pt, p, t)
|
|
|
|
|
if err == nil {
|
|
|
|
|
sc.inner = append(sc.inner, nested)
|
|
|
|
|
} else {
|
|
|
|
|
return sc, err
|
|
|
|
|
}
|
|
|
|
|
|
2014-01-29 17:50:31 -05:00
|
|
|
|
} else {
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
sc.doMarkup = currItem.typ == tLeftDelimScWithMarkup
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
|
|
|
|
cnt++
|
|
|
|
|
|
|
|
|
|
case tRightDelimScWithMarkup, tRightDelimScNoMarkup:
|
|
|
|
|
// we trust the template on this:
|
|
|
|
|
// if there's no inner, we're done
|
|
|
|
|
if !isInner {
|
|
|
|
|
return sc, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
case tScClose:
|
2015-03-02 15:23:16 -05:00
|
|
|
|
next := pt.peek()
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
if !isInner {
|
|
|
|
|
if next.typ == tError {
|
|
|
|
|
// return that error, more specific
|
|
|
|
|
continue
|
|
|
|
|
}
|
2015-02-28 13:24:30 -05:00
|
|
|
|
return sc, fmt.Errorf("Shortcode '%s' in page '%s' has no .Inner, yet a closing tag was provided", next.val, p.FullFilePath())
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
2015-03-02 15:23:16 -05:00
|
|
|
|
if next.typ == tRightDelimScWithMarkup || next.typ == tRightDelimScNoMarkup {
|
|
|
|
|
// self-closing
|
|
|
|
|
pt.consume(1)
|
|
|
|
|
} else {
|
|
|
|
|
pt.consume(2)
|
|
|
|
|
}
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
return sc, nil
|
|
|
|
|
case tText:
|
|
|
|
|
sc.inner = append(sc.inner, currItem.val)
|
|
|
|
|
case tScName:
|
|
|
|
|
sc.name = currItem.val
|
2015-04-17 18:40:54 -04:00
|
|
|
|
tmpl := getShortcodeTemplate(sc.name, t)
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
|
|
|
|
if tmpl == nil {
|
|
|
|
|
return sc, fmt.Errorf("Unable to locate template for shortcode '%s' in page %s", sc.name, p.BaseFileName())
|
|
|
|
|
}
|
2015-11-21 14:03:18 -05:00
|
|
|
|
|
2015-11-21 17:29:58 -05:00
|
|
|
|
var err error
|
|
|
|
|
isInner, err = isInnerShortcode(tmpl)
|
|
|
|
|
if err != nil {
|
|
|
|
|
return sc, fmt.Errorf("Failed to handle template for shortcode '%s' for page '%s': %s", sc.name, p.BaseFileName(), err)
|
2015-11-15 14:53:12 -05:00
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
|
|
|
|
case tScParam:
|
|
|
|
|
if !pt.isValueNext() {
|
|
|
|
|
continue
|
|
|
|
|
} else if pt.peek().typ == tScParamVal {
|
|
|
|
|
// named params
|
|
|
|
|
if sc.params == nil {
|
|
|
|
|
params := make(map[string]string)
|
|
|
|
|
params[currItem.val] = pt.next().val
|
|
|
|
|
sc.params = params
|
2014-01-29 17:50:31 -05:00
|
|
|
|
} else {
|
2015-08-07 14:08:23 -04:00
|
|
|
|
if params, ok := sc.params.(map[string]string); ok {
|
|
|
|
|
params[currItem.val] = pt.next().val
|
|
|
|
|
} else {
|
|
|
|
|
return sc, shortCodeIllegalState
|
|
|
|
|
}
|
|
|
|
|
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
|
|
|
|
} else {
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
// positional params
|
|
|
|
|
if sc.params == nil {
|
|
|
|
|
var params []string
|
|
|
|
|
params = append(params, currItem.val)
|
|
|
|
|
sc.params = params
|
2014-01-29 17:50:31 -05:00
|
|
|
|
} else {
|
2015-08-07 14:08:23 -04:00
|
|
|
|
if params, ok := sc.params.([]string); ok {
|
|
|
|
|
params = append(params, currItem.val)
|
|
|
|
|
sc.params = params
|
|
|
|
|
} else {
|
|
|
|
|
return sc, shortCodeIllegalState
|
|
|
|
|
}
|
|
|
|
|
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
|
|
|
|
case tError, tEOF:
|
|
|
|
|
// handled by caller
|
|
|
|
|
pt.backup()
|
|
|
|
|
break Loop
|
|
|
|
|
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
|
|
|
|
return sc, nil
|
|
|
|
|
}
|
|
|
|
|
|
2014-11-20 12:32:21 -05:00
|
|
|
|
func extractShortcodes(stringToParse string, p *Page, t tpl.Template) (string, map[string]shortcode, error) {
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
|
|
|
|
shortCodes := make(map[string]shortcode)
|
|
|
|
|
|
|
|
|
|
startIdx := strings.Index(stringToParse, "{{")
|
|
|
|
|
|
|
|
|
|
// short cut for docs with no shortcodes
|
|
|
|
|
if startIdx < 0 {
|
|
|
|
|
return stringToParse, shortCodes, nil
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// the parser takes a string;
|
|
|
|
|
// since this is an internal API, it could make sense to use the mutable []byte all the way, but
|
|
|
|
|
// it seems that the time isn't really spent in the byte copy operations, and the impl. gets a lot cleaner
|
|
|
|
|
pt := &pageTokens{lexer: newShortcodeLexer("parse-page", stringToParse, pos(startIdx))}
|
|
|
|
|
|
|
|
|
|
id := 1 // incremented id, will be appended onto temp. shortcode placeholders
|
2015-01-30 14:19:46 -05:00
|
|
|
|
|
|
|
|
|
result := bp.GetBuffer()
|
|
|
|
|
defer bp.PutBuffer(result)
|
|
|
|
|
//var result bytes.Buffer
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
|
|
|
|
// the parser is guaranteed to return items in proper order or fail, so …
|
|
|
|
|
// … it's safe to keep some "global" state
|
|
|
|
|
var currItem item
|
|
|
|
|
var currShortcode shortcode
|
|
|
|
|
var err error
|
|
|
|
|
|
|
|
|
|
Loop:
|
|
|
|
|
for {
|
|
|
|
|
currItem = pt.next()
|
|
|
|
|
|
|
|
|
|
switch currItem.typ {
|
|
|
|
|
case tText:
|
|
|
|
|
result.WriteString(currItem.val)
|
|
|
|
|
case tLeftDelimScWithMarkup, tLeftDelimScNoMarkup:
|
|
|
|
|
// let extractShortcode handle left delim (will do so recursively)
|
|
|
|
|
pt.backup()
|
|
|
|
|
if currShortcode, err = extractShortcode(pt, p, t); err != nil {
|
|
|
|
|
return result.String(), shortCodes, err
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if currShortcode.params == nil {
|
|
|
|
|
currShortcode.params = make([]string, 0)
|
|
|
|
|
}
|
2014-01-29 17:50:31 -05:00
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
placeHolder := createShortcodePlaceholder(id)
|
|
|
|
|
result.WriteString(placeHolder)
|
|
|
|
|
shortCodes[placeHolder] = currShortcode
|
|
|
|
|
id++
|
|
|
|
|
case tEOF:
|
|
|
|
|
break Loop
|
|
|
|
|
case tError:
|
|
|
|
|
err := fmt.Errorf("%s:%d: %s",
|
|
|
|
|
p.BaseFileName(), (p.lineNumRawContentStart() + pt.lexer.lineNum() - 1), currItem)
|
|
|
|
|
currShortcode.err = err
|
|
|
|
|
return result.String(), shortCodes, err
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
return result.String(), shortCodes, nil
|
|
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Replace prefixed shortcode tokens (HUGOSHORTCODE-1, HUGOSHORTCODE-2) with the real content.
|
2015-10-20 14:35:12 -04:00
|
|
|
|
// Note: This function will rewrite the input slice.
|
2015-06-21 07:08:30 -04:00
|
|
|
|
func replaceShortcodeTokens(source []byte, prefix string, replacements map[string]string) ([]byte, error) {
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
2015-06-21 07:08:30 -04:00
|
|
|
|
if len(replacements) == 0 {
|
|
|
|
|
return source, nil
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
|
|
|
|
|
2015-06-21 07:08:30 -04:00
|
|
|
|
sourceLen := len(source)
|
|
|
|
|
start := 0
|
|
|
|
|
|
|
|
|
|
pre := []byte("{@{@" + prefix)
|
|
|
|
|
post := []byte("@}@}")
|
|
|
|
|
pStart := []byte("<p>")
|
|
|
|
|
pEnd := []byte("</p>")
|
|
|
|
|
|
|
|
|
|
k := bytes.Index(source[start:], pre)
|
|
|
|
|
|
|
|
|
|
for k != -1 {
|
|
|
|
|
j := start + k
|
|
|
|
|
postIdx := bytes.Index(source[j:], post)
|
|
|
|
|
if postIdx < 0 {
|
|
|
|
|
// this should never happen, but let the caller decide to panic or not
|
|
|
|
|
return nil, errors.New("illegal state in content; shortcode token missing end delim")
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
end := j + postIdx + 4
|
|
|
|
|
|
|
|
|
|
newVal := []byte(replacements[string(source[j:end])])
|
|
|
|
|
|
|
|
|
|
// Issue #1148: Check for wrapping p-tags <p>
|
|
|
|
|
if j >= 3 && bytes.Equal(source[j-3:j], pStart) {
|
|
|
|
|
if (k+4) < sourceLen && bytes.Equal(source[end:end+4], pEnd) {
|
|
|
|
|
j -= 3
|
|
|
|
|
end += 4
|
2015-01-28 22:11:41 -05:00
|
|
|
|
}
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
Provide (relative) reference funcs & shortcodes.
- `.Ref` and `.RelRef` take a reference (the logical filename for a
page, including extension and/or a document fragment ID) and return
a permalink (or relative permalink) to the referenced document.
- If the reference is a page name (such as `about.md`), the page
will be discovered and the permalink will be returned: `/about/`
- If the reference is a page name with a fragment (such as
`about.md#who`), the page will be discovered and used to add the
`page.UniqueID()` to the resulting fragment and permalink:
`/about/#who:deadbeef`.
- If the reference is a fragment and `.*Ref` has been called from
a `Node` or `SiteInfo`, it will be returned as is: `#who`.
- If the reference is a fragment and `.*Ref` has been called from
a `Page`, it will be returned with the page’s unique ID:
`#who:deadbeef`.
- `.*Ref` can be called from either `Node`, `SiteInfo` (e.g.,
`Node.Site`), `Page` objects, or `ShortcodeWithPage` objects in
templates.
- `.*Ref` cannot be used in content, so two shortcodes have been
created to provide the functionality to content: `ref` and `relref`.
These are intended to be used within markup, like `[Who]({{% ref
about.md#who %}})` or `<a href="{{% ref about.md#who %}}">Who</a>`.
- There are also `ref` and `relref` template functions (used to create
the shortcodes) that expect a `Page` or `Node` object and the
reference string (e.g., `{{ relref . "about.md" }}` or `{{
"about.md" | ref . }}`). It actually looks for `.*Ref` as defined on
`Node` or `Page` objects.
- Shortcode handling had to use a *differently unique* wrapper in
`createShortcodePlaceholder` because of the way that the `ref` and
`relref` are intended to be used in content.
2014-11-24 01:15:34 -05:00
|
|
|
|
|
2015-10-20 14:35:12 -04:00
|
|
|
|
// This and other cool slice tricks: https://github.com/golang/go/wiki/SliceTricks
|
|
|
|
|
source = append(source[:j], append(newVal, source[end:]...)...)
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
|
2015-06-21 07:08:30 -04:00
|
|
|
|
k = bytes.Index(source[start:], pre)
|
2015-06-22 13:40:12 -04:00
|
|
|
|
|
2015-10-20 14:35:12 -04:00
|
|
|
|
}
|
2015-06-22 13:40:12 -04:00
|
|
|
|
|
2015-10-20 14:35:12 -04:00
|
|
|
|
return source, nil
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
}
|
2013-07-04 11:32:55 -04:00
|
|
|
|
|
2015-04-17 18:40:54 -04:00
|
|
|
|
func getShortcodeTemplate(name string, t tpl.Template) *template.Template {
|
Shortcode rewrite, take 2
This commit contains a restructuring and partial rewrite of the shortcode handling.
Prior to this commit rendering of the page content was mingled with handling of the shortcodes. This led to several oddities.
The new flow is:
1. Shortcodes are extracted from page and replaced with placeholders.
2. Shortcodes are processed and rendered
3. Page is processed
4. The placeholders are replaced with the rendered shortcodes
The handling of summaries is also made simpler by this.
This commit also introduces some other chenges:
1. distinction between shortcodes that need further processing and those who do not:
* `{{< >}}`: Typically raw HTML. Will not be processed.
* `{{% %}}`: Will be processed by the page's markup engine (Markdown or (infuture) Asciidoctor)
The above also involves a new shortcode-parser, with lexical scanning inspired by Rob Pike's talk called "Lexical Scanning in Go",
which should be easier to understand, give better error messages and perform better.
2. If you want to exclude a shortcode from being processed (for documentation etc.), the inner part of the shorcode must be commented out, i.e. `{{%/* movie 47238zzb */%}}`. See the updated shortcode section in the documentation for further examples.
The new parser supports nested shortcodes. This isn't new, but has two related design choices worth mentioning:
* The shortcodes will be rendered individually, so If both `{{< >}}` and `{{% %}}` are used in the nested hierarchy, one will be passed through the page's markdown processor, the other not.
* To avoid potential costly overhead of always looking far ahead for a possible closing tag, this implementation looks at the template itself, and is branded as a container with inner content if it contains a reference to `.Inner`
Fixes #565
Fixes #480
Fixes #461
And probably some others.
2014-10-27 16:48:30 -04:00
|
|
|
|
if x := t.Lookup("shortcodes/" + name + ".html"); x != nil {
|
|
|
|
|
return x
|
|
|
|
|
}
|
|
|
|
|
if x := t.Lookup("theme/shortcodes/" + name + ".html"); x != nil {
|
|
|
|
|
return x
|
|
|
|
|
}
|
|
|
|
|
return t.Lookup("_internal/shortcodes/" + name + ".html")
|
2013-07-04 11:32:55 -04:00
|
|
|
|
}
|
|
|
|
|
|
2015-04-17 18:40:54 -04:00
|
|
|
|
func renderShortcodeWithPage(tmpl *template.Template, data *ShortcodeWithPage) string {
|
2015-01-30 14:19:46 -05:00
|
|
|
|
buffer := bp.GetBuffer()
|
|
|
|
|
defer bp.PutBuffer(buffer)
|
|
|
|
|
|
2014-01-29 17:50:31 -05:00
|
|
|
|
err := tmpl.Execute(buffer, data)
|
|
|
|
|
if err != nil {
|
2014-03-31 13:23:34 -04:00
|
|
|
|
jww.ERROR.Println("error processing shortcode", tmpl.Name(), "\n ERR:", err)
|
|
|
|
|
jww.WARN.Println(data)
|
2014-01-29 17:50:31 -05:00
|
|
|
|
}
|
|
|
|
|
return buffer.String()
|
2013-07-04 11:32:55 -04:00
|
|
|
|
}
|