Emotes: reorganize

README.md

@@ -96,7 +96,7 @@ using the `ANONSTREAM_CONFIG` environment variable.
 
 anonstream has APIs for accessing internal state and hooking into
 internal events. They can be used by humans and other programs.  See
-[HACKING.md][/doc/HACKING.md].
+[HACKING.md](/doc/HACKING.md).
 
 ## Copying
 

anonstream/__init__.py

@@ -8,6 +8,7 @@ from collections import OrderedDict
 from quart_compress import Compress
 
 from anonstream.config import update_flask_from_toml
+from anonstream.emote import load_emote_schema
 from anonstream.quart import Quart
 from anonstream.utils.captcha import create_captcha_factory, create_captcha_signer
 from anonstream.utils.chat import precompute_emote_regex
@@ -49,10 +50,7 @@ def create_app(toml_config):
     app.allowedness = generate_blank_allowedness()
 
     # Read emote schema
-    with open(app.config['EMOTE_SCHEMA']) as fp:
+    app.emotes = load_emote_schema(app.config['EMOTE_SCHEMA'])
-        emotes = json.load(fp)
-        precompute_emote_regex(emotes)
-        app.emotes = emotes
 
     # State for tasks
     app.users_update_buffer = set()

anonstream/emote.py

@@ -0,0 +1,37 @@
+import json
+import re
+
+class BadEmoteName(Exception):
+    pass
+
+def load_emote_schema(filepath)
+    with open(filepath) as fp:
+        emotes = json.load(fp)
+        precompute_emote_regex(emotes)
+    return emotes
+
+def precompute_emote_regex(schema):
+    for emote in schema:
+        if not emote['name']:
+            raise BadEmoteName('emote names cannot be empty')
+        if re.search(r'\s', emote['name']):
+            raise BadEmoteName(
+                f'whitespace is not allowed in emote names: {emote["name"]!r}'
+            )
+        for length in (emote['width'], emote['height']):
+            if length is not None or isinstance(length, int) and length >= 0:
+                raise BadEmoteName(
+                    f'emote dimensions must be null or non-negative integers: '
+                    f'{emote["name"]!r}'
+                )
+        # If the emote name begins with a word character [a-zA-Z0-9_],
+        # match only if preceded by a non-word character or the empty
+        # string.  Similarly for the end of the emote name.
+        # Examples:
+        #  * ":joy:" matches "abc :joy:~xyz"   and   "abc:joy:xyz"
+        #  * "JoySi" matches "abc JoySi~xyz" but NOT "abcJoySiabc"
+        onset = r'(?:^|(?<=\W))' if re.fullmatch(r'\w', emote['name'][0]) else r''
+        finish = r'(?:$|(?=\W))' if re.fullmatch(r'\w', emote['name'][-1]) else r''
+        emote['regex'] = re.compile(''.join(
+            (onset, re.escape(escape(emote['name'])), finish)
+        ))

anonstream/helpers/chat.py

@@ -2,13 +2,10 @@
 # SPDX-License-Identifier: AGPL-3.0-or-later
 
 import hashlib
-from functools import lru_cache
 
-import markupsafe
+from quart import current_app
-from quart import current_app, escape, url_for, Markup
 
 CONFIG = current_app.config
-EMOTES = current_app.emotes
 
 def generate_nonce_hash(nonce):
     parts = CONFIG['SECRET_KEY'] + b'nonce-hash\0' + nonce.encode()
@@ -19,24 +16,3 @@ def get_scrollback(messages):
     if len(messages) < n:
         return messages
     return list(messages)[-n:]
-
-@lru_cache
-def get_emote_markup(emote_name, emote_file, emote_width, emote_height):
-    emote_name_markup = escape(emote_name)
-    width = '' if emote_width is None else f'width="{escape(emote_width)}" '
-    height = '' if emote_height is None else f'height="{escape(emote_height)}" '
-    return Markup(
-        f'''<img class="emote" '''
-        f'''src="{escape(url_for('static', filename=emote_file))}" '''
-        f'''{width}{height}'''
-        f'''alt="{emote_name_markup}" title="{emote_name_markup}">'''
-    )
-
-def insert_emotes(markup):
-    assert isinstance(markup, markupsafe.Markup)
-    for emote in EMOTES:
-        emote_markup = get_emote_markup(
-            emote['name'], emote['file'], emote['width'], emote['height'],
-        )
-        markup = emote['regex'].sub(emote_markup, markup)
-    return Markup(markup)

anonstream/helpers/emote.py

@@ -0,0 +1,27 @@
+import markupsafe
+from functools import lru_cache
+
+from quart import current_app, escape, url_for, Markup
+
+EMOTES = current_app.emotes
+
+@lru_cache
+def get_emote_markup(emote_name, emote_file, emote_width, emote_height):
+    emote_name_markup = escape(emote_name)
+    width = '' if emote_width is None else f'width="{escape(emote_width)}" '
+    height = '' if emote_height is None else f'height="{escape(emote_height)}" '
+    return Markup(
+        f'''<img class="emote" '''
+        f'''src="{escape(url_for('static', filename=emote_file))}" '''
+        f'''{width}{height}'''
+        f'''alt="{emote_name_markup}" title="{emote_name_markup}">'''
+    )
+
+def insert_emotes(markup):
+    assert isinstance(markup, markupsafe.Markup)
+    for emote in EMOTES:
+        emote_markup = get_emote_markup(
+            emote['name'], emote['file'], emote['width'], emote['height'],
+        )
+        markup = emote['regex'].sub(emote_markup, markup)
+    return Markup(markup)

anonstream/utils/chat.py

@@ -30,23 +30,3 @@ def get_approx_linespan(text):
     linespan = sum(map(height, text.splitlines()))
     linespan = linespan if linespan > 0 else 1
     return linespan
-
-def precompute_emote_regex(schema):
-    for emote in schema:
-        assert emote['name'], 'emote names cannot be empty'
-        assert not re.search(r'\s', emote['name']), \
-            f'whitespace is not allowed in emote names: {emote["name"]!r}'
-        for length in (emote['width'], emote['height']):
-            assert length is None or isinstance(length, int) and length >= 0, \
-                f'emote dimensions must be null or non-negative integers: {emote["name"]!r}'
-        # If the emote name begins with a word character [a-zA-Z0-9_],
-        # match only if preceded by a non-word character or the empty
-        # string.  Similarly for the end of the emote name.
-        # Examples:
-        #  * ":joy:" matches "abc :joy:~xyz"   and   "abc:joy:xyz"
-        #  * "JoySi" matches "abc JoySi~xyz" but NOT "abcJoySiabc"
-        onset = r'(?:^|(?<=\W))' if re.fullmatch(r'\w', emote['name'][0]) else r''
-        finish = r'(?:$|(?=\W))' if re.fullmatch(r'\w', emote['name'][-1]) else r''
-        emote['regex'] = re.compile(''.join(
-            (onset, re.escape(escape(emote['name'])), finish)
-        ))