add local docset generator

purs/Makefile

@@ -22,6 +22,8 @@ docs:
 	@rm -Rf generated-docs
 	@purs docs ".spago/*/*/src/**/*.purs" --format html
 
+docset: docs
+	@(cd docset; python3 ./gen-docset.py)
 clean:
 	rm -f dist/*
 

purs/docset/Info.plist.in

@@ -0,0 +1,20 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0">
+<dict>
+	<key>CFBundleIdentifier</key>
+	<string>purescript-local</string>
+	<key>CFBundleName</key>
+	<string>purescript-local</string>
+	<key>DocSetPlatformFamily</key>
+	<string>purescript</string>
+	<key>dashIndexFilePath</key>
+	<string>index.html</string>
+	<key>isDashDocset</key>
+	<true/>
+	<key>DashDocSetPlayURL</key>
+	<string>http://try.purescript.org</string>
+	<key>DashDocSetFamily</key>
+	<string>dashtoc</string>
+</dict>
+</plist>

purs/docset/gen-docset.py

@@ -0,0 +1,151 @@
+#!/usr/bin/env python3
+
+import re
+import sys
+import os
+import sqlite3
+import urllib.parse
+import requests
+import shutil
+from shutil import copyfile
+from html import unescape
+from bs4 import BeautifulSoup
+
+def fatal(msg):
+	print(msg, file=sys.stderr)
+	sys.exit(1)
+
+class URLUtilities:
+	HTML = '../generated-docs/'
+	INDEX =  HTML + 'index.html'
+
+class HTMLUtilities:
+	@staticmethod
+	def find_modules(html):
+		return re.findall(r'<li><a href="[^"]+">([^<]+)</a>', html)
+
+class Generator:
+	OUTPUT = 'purescript-local.docset'
+
+	def __init__(self):
+		self.version = None
+
+	def generate(self):
+		self.create_docset()
+		self.create_index()
+		self.save_assets()
+		modules = self.fetch_index()
+		print('Processing {} modules'.format(len(modules)))
+		for module in modules:
+			self.fetch_module(module)
+		self.db.close()
+		self.create_plist()
+		print('Done')
+
+	def fetch_index(self):
+		self.version = None
+		print('Processing module list')
+		with open(URLUtilities.INDEX, 'r') as f:
+			r = f.read()
+		html = re.sub(r'(</a></li>)</li>', r'\1', r) # fix html error
+		self.save_html(html, self.documents_path('index.html'))
+		modules = HTMLUtilities.find_modules(html)
+		return modules
+
+	@staticmethod
+	def create_docset():
+		path = Generator.OUTPUT
+		if os.path.exists(path):
+			shutil.rmtree(path)
+		os.makedirs(Generator.documents_path())
+
+	def fetch_module(self, module):
+		print('Processing module {}'.format(module))
+		moduleFile = urllib.parse.quote(module, '') + '.html'
+		with open('{}/{}'.format(URLUtilities.HTML, moduleFile), 'r') as f:
+			r = f.read()
+		html = self.save_html(r, self.documents_path(moduleFile))
+		self.cursor.execute(
+			'INSERT OR IGNORE INTO searchIndex(name, type, path) VALUES (?,?,?);',
+			[module, 'Module', moduleFile])
+		self.db.commit()
+		return r
+
+	@staticmethod
+	def documents_path(*paths):
+		return os.path.join(Generator.OUTPUT, 'Contents/Resources/Documents', *paths)
+
+	def save_html(self, html, path):
+		prefix = r''
+		soup = BeautifulSoup(html, 'html.parser')
+		# remove google font
+		soup.find('link', href=re.compile(r'^https://fonts\.googleapis\.com/.*')).decompose()
+		# remove top banner
+		soup.find('div', class_='top-banner').decompose()
+		aside = soup.find('div', class_='col--aside')
+		if(aside):
+			aside.decompose()
+		# find anchors
+		tlds = soup.find_all('div', class_='decl')
+		for tld in tlds:
+			self.process_decl(path, tld, soup)
+		with open(path, 'w') as f:
+			f.write(str(soup))
+
+	def process_decl(self, path, decl, soup, type_hint = None):
+		type_, name = decl.get('id').split(':', 1)
+		name = unescape(name)
+		if type_hint:
+			type_ = type_hint
+		else:
+			type_ = self.convert_type(type_)
+			signature = decl.find('pre', class_='decl__signature')
+			if signature:
+				if signature.code.find() == signature.code.find('span', class_='keyword', text='class'):
+					type_ = 'Class'
+		anchor_toc = '//apple_ref/cpp/{}/{}'.format(urllib.parse.quote(type_, ''), urllib.parse.quote(name, ''))
+		self.cursor.execute(
+			'INSERT OR IGNORE INTO searchIndex(name, type, path) VALUES (?,?,?);',
+			[name, type_, '{}#{}'.format(os.path.relpath(path, self.documents_path()), anchor_toc)])
+		a = soup.new_tag('a', attrs={ 'name': anchor_toc, 'class': 'dashAnchor' })
+		decl.insert(0, a)
+		if type_ == 'Class':
+			members_lbl = decl.find('h4', text='Members')
+			if members_lbl:
+				for member in members_lbl.find_next_sibling().find_all('li', recursive=False):
+					self.process_decl(path, member, soup, 'Function')
+		elif type_ == 'Type':
+			ctors_lbl = decl.find('h4', text='Constructors')
+			if ctors_lbl:
+				for ctor in ctors_lbl.find_next_sibling().find_all('li', recursive=False):
+					self.process_decl(path, ctor, soup, 'Constructor')
+
+	def save_assets(self):
+		copyfile('favicon-16x16.png', self.documents_path('../../../icon.png'))
+		copyfile('favicon-32x32.png', self.documents_path('../../../icon@2x.png'))
+
+	@staticmethod
+	def create_plist():
+		with open('Info.plist.in', 'r') as f:
+			plist = f.read()
+		with open(Generator.documents_path('../../Info.plist'), 'w') as f:
+			f.write(plist)
+
+	@staticmethod
+	def convert_type(t):
+		TABLE = {
+			't': 'Type',
+			'v': 'Function',
+			'k': 'Kind',
+		}
+		return TABLE[t] if t in TABLE else t
+
+	def create_index(self):
+		self.db = sqlite3.connect(self.documents_path('../docSet.dsidx'))
+		self.cursor = self.db.cursor()
+		self.cursor.execute('CREATE TABLE searchIndex(id INTEGER PRIMARY KEY, name TEXT, type TEXT, path TEXT);')
+		self.cursor.execute('CREATE UNIQUE INDEX anchor ON searchIndex (name, type, path);')
+
+if __name__ == '__main__':
+	gen = Generator()
+	gen.generate()