包详细信息

@unified-latex/unified-latex-util-replace

siefkenj34.8kMIT1.8.3

Functions for modifying a unified-latex AST

pegjs, latex, parser, prettier

自述文件

unified-latex-util-replace

What is this?

Functions to help modify a unified-latex Abstract Syntax Tree (AST).

When should I use this?

If you want to recursively replace particular AST nodes.

Install

npm install @unified-latex/unified-latex-util-replace

This package contains both esm and commonjs exports. To explicitly access the esm export, import the .js file. To explicitly access the commonjs export, import the .cjs file.

Plugins

unifiedLatexReplaceStreamingCommands

Unified plugin to replace all found streaming commands with their argument-style equivalents. This only applies to sections of the tree with no math ancestor.

Usage

unified().use(unifiedLatexReplaceStreamingCommands[, options])

options

PluginOptions

Type

Plugin<PluginOptions[], Ast.Root, Ast.Root>

function unifiedLatexReplaceStreamingCommands(
  options: PluginOptions
): (tree: Ast.Root) => void;

where

type PluginOptions = {
  replacers: Record<
    string,
    (content: Ast.Node[], streamingCommand: Ast.Macro) => Ast.Node | Ast.Node[]
  >;
};

Functions

firstSignificantNode(nodes, parbreaksAreInsignificant)

Returns the first non-whitespace/non-comment node in nodes. If there is no such node, null is returned.

function firstSignificantNode(
  nodes: Ast.Node[],
  parbreaksAreInsignificant: Boolean
): Ast.Node;

Parameters

Param Type
nodes Ast.Node[]
parbreaksAreInsignificant Boolean

firstSignificantNodeIndex(nodes, parbreaksAreInsignificant)

Returns the index of the first non-whitespace/non-comment node in nodes. If there is no such node, null is returned.

function firstSignificantNodeIndex(
  nodes: Ast.Node[],
  parbreaksAreInsignificant: Boolean
): number;

Parameters

Param Type
nodes Ast.Node[]
parbreaksAreInsignificant Boolean

lastSignificantNode(nodes, parbreaksAreInsignificant)

Returns the last non-whitespace/non-comment node in nodes. If there is no such node, null is returned.

function lastSignificantNode(
  nodes: Ast.Node[],
  parbreaksAreInsignificant: Boolean
): Ast.Node;

Parameters

Param Type
nodes Ast.Node[]
parbreaksAreInsignificant Boolean

lastSignificantNodeIndex(nodes, parbreaksAreInsignificant)

Returns the index of the last non-whitespace/non-comment node in nodes. If there is no such node, null is returned.

function lastSignificantNodeIndex(
  nodes: Ast.Node[],
  parbreaksAreInsignificant: Boolean
): number;

Parameters

Param Type
nodes Ast.Node[]
parbreaksAreInsignificant Boolean

replaceNode(ast, visitor)

Recursively replace nodes in ast. The visitor function is called on each node. If visitor returns a node or an array of nodes, those nodes replace the node passed to visitor. If null is returned, the node is deleted. If undefined is returned, no replacement happens.

function replaceNode(
  ast: Ast.Ast,
  visitor: (
    node: Ast.Node | Ast.Argument,
    info: VisitInfo
  ) =>
    | Ast.Node
    | Ast.Argument
    | (Ast.Node | Ast.Argument)[]
    | null
    | undefined
    | void
): void;

Parameters

Param Type
ast Ast.Ast
visitor Omitted

replaceNodeDuringVisit(replacement, info)

Replaces the current node with replacement. It is assumed that the current node is in an array that is a child of a parent element. If this is not the case, the function will error.

function replaceNodeDuringVisit(
  replacement: Ast.Node | Ast.Argument | (Ast.Node | Ast.Argument)[],
  info: VisitInfo
): void;

Parameters

Param Type
replacement Omitted
info VisitInfo

where

type VisitInfo = {
  /**
   * If the element was accessed via an attribute, the attribute key is specified.
   */
  readonly key: string | undefined;
  /**
   * If the element was accessed in an array, the index is specified.
   */
  readonly index: number | undefined;
  /**
   * A list of ancestor nodes, `[parent, grandparent, great-grandparent, ...]`
   */
  readonly parents: (Ast.Node | Ast.Argument)[];
  /**
   * If the element was accessed in an array, the array that it is part of.
   */
  readonly containingArray: (Ast.Node | Ast.Argument)[] | undefined;
  /**
   * The LaTeX context of the current match.
   */
  readonly context: VisitorContext;
};

replaceStreamingCommand(ast, isStreamingCommand, replacer, options)

Given a group or a node array, look for streaming commands (e.g., \bfseries) and replace them with the specified macro. The "arguments" of the streaming command are passed to replacer and the return value of replacer is inserted into the stream.

By default, this command will split at parbreaks (since commands like \textbf{...} do not accept parbreaks in their contents) and callreplacer` multiple times, once per paragraph.

Commands are also split at environments and at any macros listed in macrosThatBreakPars.

function replaceStreamingCommand(
  ast: Ast.Group | Ast.Node[],
  isStreamingCommand: (node: any) => node is Ast.Macro,
  replacer: (
    content: Ast.Node[],
    streamingCommand: Ast.Macro
  ) => Ast.Node | Ast.Node[],
  options: {
    macrosThatBreakPars?: string[];
    environmentsThatDontBreakPars?: string[];
  }
): Ast.Node[];

Parameters

Param Type
ast `Ast.Group \ Ast.Node[]`
isStreamingCommand Omitted
replacer Omitted
options Omitted

replaceStreamingCommandInGroup(group, isStreamingCommand, replacer, options)

Process streaming commands in a group. If needed, "escape" the group. For example, {\bfseries xx} -> \textbf{xx}, but {foo \bfseries xx} -> {foo \textbf{xx}}.

function replaceStreamingCommandInGroup(
  group: Ast.Group,
  isStreamingCommand: (node: any) => node is Ast.Macro,
  replacer: (
    content: Ast.Node[],
    streamingCommand: Ast.Macro
  ) => Ast.Node | Ast.Node[],
  options: {
    macrosThatBreakPars?: string[];
    environmentsThatDontBreakPars?: string[];
  }
): Ast.Node[];

Parameters

Param Type
group Ast.Group
isStreamingCommand Omitted
replacer Omitted
options Omitted

更新日志

unified-latex Changelog

v1.8.3

  • Support \ref in PreTeXt conversion
  • Better use of UnifiedJS to parse but not print LaTeX
  • Support for \verb, \textsuperscript, \textsubscript, \sout, and \" i in HTML conversion

v1.8.2

  • Upgraded dependencies

v1.8.1

  • Changed Peggy to implement a caching parser to prevent large slowdown on some files.

v1.8.0

  • Added initial PreTeXt conversion support
  • Upgraded deps
  • Added amsart macros
  • Consume the whitespace after special character macros when expanding ligatures. For example \o y produces øy instead of ø y
  • Fix signatures of \hyphenation

v1.7.1

  • Types fix for @unified-latex/unified-latex-types
  • Fixed AST when expanding \sysdelim macros for rendering \systeme{} macros with KaTeX

v1.7.0

  • Switch build system to vite. Should result in smaller bundles.
  • Save default arguments when parsing if the macro signature specifies them e.g. {signature: "O{foo}"}. The defaults are substituted in when expanding the macros with the optional arguments omitted.
  • Preserve position information when comments are modified. (Sometimes, during a parse, but never during a parseMinimal, comments are modified to remove leading whitespace. Previously, modified comments would have their position information deleted. Position information is now preserved.)

v1.6.1

  • Pass VisitInfo as an additional argument ot macroReplacers and environmentReplacers in unifiedLatexToHast.
  • Allow skipping of HTML validation in unifiedLatexToHast.
  • The minted environment parses its contents as a verbatim.

v1.6.0

  • Embellishment tokens are now supported in macro signatures. E.g., a xxx: {signature: "e{^_}"} will allow \xxx_{foo}^{bar} and \xxx^{foo}_{bar} to parse correctly.
  • Stop tokens can now be regular string characters. For example xxx: {signature: "ua"} will allow \xxx YYYaBBB to consume YYY leaving BBB unconsumed.
  • Break after \\ macro when pretty printing (Issue #59)
  • [DEVELOPMENT] Added tsconfig.json files to each test/ folder for more granular control of the typescript settings.

v1.5.0

  • HTML conversion: vspace and hspace now give the amount in a data-amount attribute.
  • HTML conversion: unknown macros now have their arguments wrapped in spans instead of appearing as formatted LaTeX code.
  • Add basic Markdown conversion support.

v1.4.2

  • Avoid slowdown when paring incomplete environments (e.g. \newcommand{\x}{\begin{x}}). This is accomplished by enabling caching in PEGjs.
  • Added " ligature and \paragraph and \subparagraph to HTML conversion.

v1.4.1

  • Many more ligatures added to the HTML converter.
  • Fixed issue #40 where the optional argument to \\ was being parsed even if preceded by a space. (E.g., \\[10pt] and \\ [10pt]) were parsed the same. Not allowing the space should more closely match expected behavior.
  • Bump Prettier to v2.8.8

v1.4.0

  • Better CJS support (now unified is compiled in rather than left as an external dependency. This is needed because unified is ESM-only).
  • minted and listings environments now accept optional arguments and parse their contents verbatim. This makes them much more efficient.

v1.3.0

  • Initial support for parsing and pretty-printing of tikz environments.
  • Added support for xparse u-type arguments.
  • Can now pass an argumentParser attribute for custom argument parsing (instead of relying on an xparse signature)