10 files changed, 209 insertions(+), 48 deletions(-)

M .gitignore
A gloss/__main__.py
A gloss/markup/__init__.py
A gloss/markup/markup
R {markup => gloss/markup/markup}.c
R {gloss => gloss/modes/glossary}.py
A run
A testsrc/.gloss
A testsrc/a.html
A testsrc/b.html

@@ 1,1 1,2 @@
 /markup
+__pycache__

@@ 0,0 1,103 @@
+#!/bin/env python3
+
+from sys import argv
+import yaml
+from pathlib import Path
+from typing import List, Tuple, Callable
+from dataclasses import dataclass
+
+from . import markup
+
+from .modes import glossary
+
+usage = f'''
+usage: {argv[0]} <SRCDIR> <OUTDIR>
+
+recurses into SRCDIR, parsing each .gloss file as YAML and running the
+requested parser. if a directory does not contain a '.gloss' file, or
+said file's 'mode' key is missing, the mode defaults to 'copy'.
+
+.gloss may include the following fields:
+	mode:	sets the mode. may be one of 'copy', 'ignore', or
+		'gloss'. if mode is 'ignore', the directory and all
+		subdirectories will be skipped.
+	out:	sets the output directory, relative to OUTDIR. if
+		omitted, it defaults to the current path relative to
+		SRCDIR.
+'''[1:]
+
+def die(why, code=1):
+	print(why, file=stderr)
+	exit(code)
+
+
+### MODES ###
+
+def copymode(srcdir, outdir):
+	for srcfile in srcdir.iterdir():
+		if not srcfile.is_file() or srcfile.name == '.gloss':
+			continue
+		outfile = outdir + srcfile.name
+		outfile.write_bytes(srcfile.read_bytes())
+
+modes = {
+	'copy': copymode,
+	'gloss': glossary.gloss
+}
+
+
+### FS WALK ###
+
+@dataclass
+class SrcDir:
+	srcdir: Path
+	outdir: Path
+	process: Callable[[Path, Path], None]
+
+def get_dirs(srcdir, outdir) -> List[SrcDir]:
+	global modes
+
+	dirs = [srcdir]
+	ret = []
+	for adir in dirs:
+		if not adir.is_dir():
+			continue
+
+		glossfile = Path(adir, '.gloss')
+		out = mode = None
+		if glossfile.is_file():
+			parsed = yaml.safe_load(open(glossfile, 'rt'))
+			out = parsed.get('out', None)
+			mode = parsed.get('mode', None)
+		
+		if out is None:
+			out = adir.relative_to(srcdir)
+		if mode is None:
+			mode = 'copy'
+		
+		if mode == 'ignore':
+			continue
+		
+		if mode not in modes:
+			die(f'fatal: unknown mode in "{glossfile}": "{mode}"')
+		
+		ret.append(SrcDir(adir, Path(outdir, out), modes[mode]))
+		dirs.extend(adir.iterdir())
+	
+	return ret
+
+
+if __name__ == '__main__':
+	argv = argv[1:]
+	if len(argv) != 2 or argv[0][0] == '-' or argv[1][0] == '-':
+		die(usage)
+
+	srcdir = Path(argv[0]).absolute()
+	outdir = Path(argv[1]).absolute()
+	
+	steps = get_dirs(srcdir, outdir)
+	print(steps)
+	for step in steps:
+		if not step.outdir.exists():
+			step.outdir.mkdir()
+		step.process(step.srcdir, step.outdir)

@@ 0,0 1,46 @@
+import html
+from pathlib import Path
+from subprocess import Popen, PIPE, run as runcmd
+
+class Markup:
+	def __init__(self, where, *, cfile='markup.c', bfile='markup'):
+		self.cfile = Path(where, cfile)
+		self.bfile = Path(where, bfile)
+
+		if not self.bfile.exists() or self.cfile.stat().st_mtime > self.bfile.stat().st_mtime:
+			print("recompiling markup subsystem...")
+			runcmd(['cc', str(self.cfile), '-o', str(self.bfile), '-Wall'], check=True)
+			print("recompiled!")
+		
+		self.proc = Popen([str(self.bfile), 'convert'], stdin=PIPE, stdout=PIPE, text=True)
+	
+	def process(self, text):
+		self.proc.stdin.write(text+"\n")
+		self.proc.stdin.flush()
+
+		segments = []
+		while (line := self.proc.stdout.readline()) not in ('NEXT\n', ''):
+			ty = line[0:4]
+
+			length = int(line[5:9])
+			ltext = line[12:12+length]
+
+			if ty == 'HTML':
+				segments.append((ltext,))
+			elif ty == 'IESC':
+				segments.append((html.escape(ltext),))
+			elif ty == 'TEXT':
+				segments.append(ltext)
+			else:
+				print(f'read in unknown type "{ty}" from markup subprocess', file=stderr)
+
+		return segments
+
+markup = None
+
+def get():
+	global markup
+	if markup is None:
+		markup = Markup(Path(__file__).parent)
+	
+	return markup

@@ 262,7 262,6 @@ segmentize(size_t length, const char *source)
 		}
 	}
 	if(in_italic){
-		PLINE();
 		tail->next = italic_head->next;
 		if(tail->next != NULL)
 			tail = italic_tail;

@@ 8,8 8,8 @@ from glob import glob
 from typing import List, Optional, Union, Dict, Set, Tuple
 from dataclasses import dataclass
 from sys import argv, stderr, exit
-from pathlib import Path
-from subprocess import Popen, PIPE, run as runcmd
+
+from .. import markup
 
 usage = f'''
 usage: {argv[0]} <SRCDIR> <OUTDIR>


@@ 116,42 116,7 @@ def first_pass(slug, fp):
 	elif block_type is not None:
 		blocks.append(Block(block_type, block_text, block_meta))
 	
-	return GlsFile(slug, title, [*names], blocks, see_also)
-
-class Markup:
-	def __init__(self, where, *, cfile='markup.c', bfile='markup'):
-		self.cfile = Path(where, cfile)
-		self.bfile = Path(where, bfile)
-
-		if not self.bfile.exists() or self.cfile.stat().st_mtime > self.bfile.stat().st_mtime:
-			print("recompiling markup subsystem...")
-			runcmd(['cc', str(self.cfile), '-o', str(self.bfile), '-Wall'], check=True)
-			print("recompiled!")
-		
-		self.proc = Popen([str(self.bfile), 'convert'], stdin=PIPE, stdout=PIPE, text=True)
-	
-	def process(self, text):
-		self.proc.stdin.write(text+"\n")
-		self.proc.stdin.flush()
-
-		segments = []
-		while (line := self.proc.stdout.readline()) not in ('NEXT\n', ''):
-			ty = line[0:4]
-
-			length = int(line[5:9])
-			ltext = line[12:12+length]
-
-			if ty == 'HTML':
-				segments.append((ltext,))
-			elif ty == 'IESC':
-				segments.append((html.escape(ltext),))
-			elif ty == 'TEXT':
-				segments.append(ltext)
-			else:
-				print(f'read in unknown type "{ty}" from markup subprocess', file=stderr)
-
-		return segments
-			
+	return GlsFile(slug, title, [*names], blocks, see_also)			
 
 quote_pat = re.compile("((?:(?!@@)(?!//).)*)(?:@@((?:(?!//).)+))?(?://(.+))?")
 


@@ 264,14 229,7 @@ def gen_inner_html(fmt, file, idx):
 
 ### ENTRYPOINT ###
 
-if __name__ == '__main__':
-	argv = argv[1:]
-	if len(argv) != 2 or argv[0][0] == '-' or argv[1][0] == '-':
-		die(usage)
-
-	srcdir = argv[0]
-	outdir = argv[1]
-
+def gloss(srcdir, outdir):
 	try:
 		with open('{srcdir}/template.html', 'rt') as fp:
 			template = fp.read()


@@ 294,7 252,7 @@ if __name__ == '__main__':
 </html>
 '''
 
-	fmt = Markup(Path(__file__).parent)
+	fmt = markup.get()
 
 	files = []
 	for fn in glob('*.gls', root_dir=srcdir):


@@ 315,3 273,10 @@ if __name__ == '__main__':
 				'modtime': datetime.fromtimestamp(stat(f'{srcdir}/{file.slug}.gls').st_mtime)
 			}
 			fp.write(template.format(**ctx))
+
+if __name__ == '__main__':
+	argv = argv[1:]
+	if len(argv) != 2 or argv[0][0] == '-' or argv[1][0] == '-':
+		die(usage)
+
+	gloss(argv[0], argv[1])

@@ 0,0 1,3 @@
+#!/bin/sh
+
+PYTHONPATH="$(dirname "$0")" python -m gloss $@

@@ 0,0 1,1 @@
+mode: gloss

@@ 0,0 1,24 @@
+<!doctype html>
+<html>
+<head>
+	<meta encoding='utf-8'>
+	<title>The Letter A</title>
+</head>
+<body>
+<main>
+<h1>The Letter A</h1>
+<p>this is a paragraph block. these lines will be folded into one string and ultimately rendered roughly the same in the browser.</p>
+<div>
+	<blockquote cite="https://aleteoryx.me">this is a quote block</blockquote>
+	<p>&mdash; <cite><a href="https://aleteoryx.me">aleteoryx</a></cite>, 2025</p>
+</div>
+<p>\&lt;http://example.com|example page&gt;</p>
+</main>
+<hr>
+<h3>See also:</h3>
+<ul>
+	<li><a href="b.html">The Letter B</a></li>
+</ul>
+<footer>File last modified Fri, 2025-15-51 00:51:39 </footer>
+</body>
+</html>

@@ 0,0 1,19 @@
+<!doctype html>
+<html>
+<head>
+	<meta encoding='utf-8'>
+	<title>The Letter B</title>
+</head>
+<body>
+<main>
+<h1>The Letter B</h1>
+<p>preceeded by <a href="a.html">the letter a</a>.</p>
+</main>
+<hr>
+<h3>See also:</h3>
+<ul>
+	<li><a href="a.html">The Letter A</a></li>
+</ul>
+<footer>File last modified Thu, 2025-14-40 18:40:52 </footer>
+</body>
+</html>