parsing html xml parsing expression grammar

By Xah Lee. Date: 2025-09-29.

2025-09-30

parsing expression grammar

;; -*- coding: utf-8; lexical-binding: t; -*-

(require 'peg)

;; (define-peg-ruleset 'number-grammar
;;         '((number sign digit (* digit))
;;           digit  ;; A reference to the definition above.
;;           (sign (or "+" "-" ""))))

(peg-parse
  (number sign digit (* digit))
  (sign   (or "+" "-" ""))
  (digit  [0-9]))

;; -206

;; -*- coding: utf-8; lexical-binding: t; -*-

(require 'peg)

(let ((xhtml
"<section class=\"word88\">
<h3 class=\"wd\">sanctimonious</h3>
<div class=\"ex\">
<div class=\"bdy\">
Trump's victory was fueled by fatigue with <b class=\"w\">sanctimonious</b> identity politics, which became state religion under Democratic administrations.</div>
<div class=\"src\">
Camille Paglia</div>
</div>

</section>"
)
      xtree
      (xoutbuf (generate-new-buffer "*peg parse out*")))

  ;; code

  (pop-to-buffer xoutbuf))

2025-09-25

xtodo

parsing html xml

think about parsing html problem.
e.g. i often need to redo html tags, e.g. vocabulary word88
using emacs lisp with regex is very tedius, cuz cannot parse nesting.
look into PEG parsing expression grammar. or html parser.
sample page
SAT Words

2025-09-29

xtodo

libxml-parse-html-region
libxml-parse-xml-region
shr-insert-document

(libxml-available-p)
;; t

(let ((xhtml "<!DOCTYPE html>
<html>
<head>
<meta charset=\"utf-8\" />
<meta name=viewport content=\"width=device-width, initial-scale=1\" />

<title>untitled</title>
</head>
<body>

<main>

<h1>untitled</h1>

<p>
some
</p>

</main>

</body>
</html>")
      xtree
      (xoutbuf (generate-new-buffer "html parse out*")))

  (setq xtree
        (with-temp-buffer
          (insert xhtml)
          (libxml-parse-html-region (point-min) (point-max))))

  (with-current-buffer xoutbuf (shr-insert-document xtree))
  (pop-to-buffer xoutbuf))

(libxml-available-p)
;; t

(let ((xhtml "<html>
<head>
<meta charset=\"utf-8\" />

<title>untitled</title>
</head>
<body>

<main>

<h1>untitled</h1>

<p>
some
</p>

</main>

</body>
</html>")
      xtree
      (xoutbuf (generate-new-buffer "html parse out*")))

  (setq xtree
        (with-temp-buffer
          (insert xhtml)
          (libxml-parse-xml-region (point-min) (point-max))))

  (with-current-buffer xoutbuf (shr-insert-document xtree))
  (pop-to-buffer xoutbuf))

2025-09-28

xtodo

when done study pasing html reformat my site citation.

〔<cite>...</cite> <time>...</time> @ <a ...>...</a>〕
<cite>...</cite> <time>...</time> @ <a ...>...</a>