#css-selectors #html #html-content #html-css #selector #command #pipeline

bin+lib html-streaming-editor

UNIX-tool like streaming editor for HTML content

15 releases (8 breaking)

0.8.0 May 13, 2023
0.7.1 Oct 19, 2022
0.6.0 Oct 13, 2022
0.0.1 Jul 30, 2022

#102 in Text editors

MIT license

205KB
4.5K SLoC

HTML Streaming Editor

License: MIT GitHub Workflow Status docs.rs Crates.io

Run (simple) manipulations on HTML files, like extracting parts. Use CSS selectors to define which parts of the HTML to operator on, use different commands in pipes to perform the desired operations.

Syntax

The basic syntax is:

COMMAND{ SELECTOR } | COMMAND{ SELECTOR }

Some COMMAND use sub-pipelines. There are two kind of COMMANDS with this:

  • "iterate"/"forEach": For each (sub) node matching the inner selector the sub-pipeline is processed, but the elements themselves are not changed

    COMMAND{ SELECTOR ↦ COMMAND{ SELECTOR } | COMMAND { SELECTOR } }

    COMMAND{ SELECTOR => COMMAND{ SELECTOR } | COMMAND { SELECTOR } }

  • "map"/"replace": For each (sub) node matching the inner selector the sub-pipeline is processed, and the pipelines result is used instead of the element

    COMMAND{ SELECTOR ↤ COMMAND{ SELECTOR } | COMMAND { SELECTOR } }

    COMMAND{ SELECTOR <= COMMAND{ SELECTOR } | COMMAND { SELECTOR } }

The SELECTOR is a CSS selector.

Pipeline Types

There are three types of pipelines:

  • element processing pipeline: The default. You have some input HTML which you run through the pipeline
  • element creating sub-pipeline: special sub-pipeline wherever a commands adds one or more elements into the HTML tree (or into a different place of said tree)
  • string value creating sub-pipeline: special sub-pipeline wherever a commands set a string value (text content, comment, attribute value)

Commands

Currently supported element processing commands:

  • EXTRACT-ELEMENT: remove everything not matching the CSS selector (alias: ONLY)
  • REMOVE-ELEMENT: remove everything matching the CSS selector (alias: WITHOUT)
  • FOR-EACH: run a sub-pipeline on all sub-elements matching a CSS selector but return the previously selected elements (alias: WITH)
  • REPLACE-ELEMENT: replace all elements matching a CSS selector with new elements (alias: MAP)
  • CLEAR-ATTR: removes a given attribute from the previously selected elements
  • SET-ATTR: Sets a given attribute to a specified value
  • CLEAR-CONTENT: clears all children from the previously selected elements
  • SET-TEXT-CONTENT: removes previous children and replaces it with exactly one given text child
  • APPEND-TEXT-CONTENT: appends a new text child
  • PREPEND-TEXT-CONTENT: prepends a new text child
  • APPEND-COMMENT: appends a new comment child
  • PREPEND-COMMENT: prepends a new comment child
  • APPEND-ELEMENT: appends a new tag/element child
  • PREPEND-ELEMENT: prepends a new tag/element child

Currently supported element creating commands:

  • CREATE-ELEMENT: creates a new, empty element, mainly in combination with APPEND-ELEMENT, PREPEND-ELEMENT or REPLACE-ELEMENT (alias: NEW)
  • LOAD-FILE: reads a DOM from a different file, mainly in combination with APPEND-ELEMENT, PREPEND-ELEMENT or REPLACE-ELEMENT (alias: SOURCE)
  • QUERY-REPLACED: returns children matching the CSS selector of those elements meant to be replaced, only combination with or REPLACE-ELEMENT (alias: KEEP)

Currently supported string-value creating commands:

  • USE-ELEMENT: returns the currently selected element for a sub-pipeline, mainly in combination with "string value producing pipelines" (alias: THIS)
  • USE-PARENT: returns the parent of the currently selected element for a sub-pipeline, mainly in combination with "string value producing pipelines" (alias: PARENT)
  • QUERY-ELEMENT: runs a query on the currently selected element for a sub-pipeline, without detaching target element from HTML tree unlike EXTRACT-ELEMENT
  • QUERY-PARENT: runs a query on the parent of the currently selected element for a sub-pipeline
  • QUERY-ROOT: runs a query on the root of the currently selected element for a sub-pipeline
  • GET-ATTR: returns the value of an attribute of the currently selected element for a string-value producing pipelines
  • GET-TEXT-CONTENT: returns the text content of the currently selected element for a string-value producing pipelines
  • REGEX-REPLACE: runs a RegEx-based value replacements on the current string value of the pipeline
  • TO-LOWER: all-lower the current string value of the pipeline
  • TO-UPPER: all-caps the current string value of the pipeline
  • ADD-PREFIX: add a given string to the beginning of the string value of the pipeline
  • ADD-SUFFIX: add a given string to the end of the string value of the pipeline

Binary

The binary is called hse and supports following options:

USAGE:
    hse [OPTIONS] <PIPELINE>

ARGS:
    <PIPELINE>  Single string with the command pipeline to perform. If it starts with an @ the rest is treated as file name to read the pipeline definition from

OPTIONS:
    -h, --help               Print help information
    -i, --input <input>      File name of the Input. `-` for stdin (default)
    -o, --output <output>    File name of the Output. `-` for stdout (default)
    -V, --version            Print version information

Example

# fetches all elements with CSS class "content" inside a <main> element
hse -i index.html 'ONLY{main .content}'

# fetches the `<main>` or element with CSS class `main`, but without any `<script>` defined inside
hse -i index.html 'ONLY{main, .main} | WITHOUT{script}'

# replaces all elements with `placeholder` class with the <div class="content"> from a second HTML file 
hse -i index.html 'MAP{.placeholder ↤ SOURCE{"other.html"} | ONLY{div.content} }'

# add a new <meta name="version" value=""> element to <head> with git version info 
hse -i index.html "WITH{head ↦ APPEND-ELEMENT{ NEW{meta} | SET-ATTR{name ↤ 'version'} | SET-ATTR{content ↤ '`git describe --tags`'}  } }"

# add a new comment to <body> with git version info
hse -i index.html "WITH{body ↦ APPEND-COMMENT{'`git describe --tags`'}}"

# add an RDF <meta name="dc:title"> with same content as <title>
hse -i input.html "WITH{head ↦ APPEND-ELEMENT{ NEW{meta} | SET-ATTR{name ↤ 'dc:title' } } | WITH{meta[name='dc:title'] ↦ SET-ATTR{content ↤ QUERY-PARENT{title} | GET-TEXT-CONTENT } } }"

# replace non-word characters with an underscore in an attribute
hse -i index.html "EXTRACT-ELEMENT{#target} | SET-ATTR{data-test ↤ USE-ELEMENT | GET-ATTR{data-test} | REGEX-REPLACE{'\\W' ↤ '_'} }"

# run the pipeline defined in file `file.hsp` on content of `index.html`
hse -i index.html @file.hsp

Dependencies

~6.5–9.5MB
~177K SLoC