more reliable and escapable formatting; add [dim]

2 files changed, 40 insertions, 32 deletions

diff --git a/clients/urwid/main.py b/clients/urwid/main.py
index 9a43e43..4b0e934 100644
--- a/clients/urwid/main.py
+++ b/clients/urwid/main.py
@@ -83,24 +83,27 @@ format_help = [
 
     "[red: Whitespace]",
 
+    "Single line breaks in the body join into eachother to form sentences, "
+    "putting a space where the break was. This works like html. When you want "
+    "to split it off into a paragraph, **use two or more line breaks.**",
+
     "When you're composing, it is desirable to introduce linebreaks into the "
     "body to keep it from overflowing the screen. However, you __dont__ want "
     "that same spacing to bleed over to other people's screens, because clients "
-    "will wrap the text themselves.",
+    "will wrap the text themselves according to user preferences or implementation "
+    "details. Thats why it works like this.",
 
-    "Single line breaks in the body join into eachother to form sentences, "
-    "putting a space where the break was. This works like html. When you want "
-    "to split it off into a paragraph, **use two line breaks.**",
 
     "[red: Colors, Bold, Underline & Expressions]",
 
     "You can use [rainbow: rainbow], [red: red], [yellow: yellow], [green: green], "
     "[blue: blue], [cyan: cyan], [magenta: and magenta], **bold**, and __underline__ "
-    "inside of your posts. **bold\nworks like this**, __and\nunderlines like this__. "
-    "The symbolic, markdown form of these directives does NOT allow escaping, and "
-    "can only apply to up to 20 characters on the same line. They are best used on short "
-    "phrases. However, you can use a different syntax for it, which is also required to use "
-    "colors: these expressions \[bold: look like this] and are much more reliable. "
+    "inside of your posts. \**bold works like this\**, \__and underlines like this\__. "
+    "You can escape these expressions \\\**like this\\\**. They can span up to the full width "
+    "of the same line. They are best used on shorter phrases. "
+    "However, you can use a different syntax for it, which is also required to use "
+    "colors: these expressions \[bold: look like this] and have less restrictions.",
+
     "The colon and the space following it are important. When you use these "
     "expressions, the __first__ space is not part of the content, but any characters, "
     "including spaces, that follow it are included in the body. The formatting will "
@@ -116,8 +119,8 @@ format_help = [
 
     "The following directives may be used in this form: red, yellow, green, blue, cyan, "
     "magenta, bold, underline, and rainbow. Nesting expressions into eachother will "
-    "override the parent directives until it closes. Thus, nesting is valid but doesn't produce "
-    "layered results.",
+    "override the parent directives until the innermost expression closes. Thus, nesting "
+    "is valid but doesn't produce layered results on the command line client.",
 
     "[red: Quotes & Greentext]",
 
@@ -126,7 +129,9 @@ format_help = [
     "this violates the sentence structure outlined in the **Whitespace** section above, "
     "so you may introduce >greentext without splicing into seperate paragraphs. The '>' "
     "must be the first character on the line with no whitespace before it.\n>it looks like this\n"
-    "and the paragraph doesnt have to break on either side.",
+    "and the paragraph doesnt have to break on either side. The formatter is smart enough to "
+    "differentiate between >>greentext with multiple arrows and numeric quotes (outlined below) "
+    "given that the text doesn't start with any numbers.",
 
     "When using numeric quotes, they are highlighted and the author's name will show "
     "next to them in the thread. You can press enter when focused on a message to view "
@@ -949,7 +954,7 @@ class App(object):
             widget, app.loop.widget,
             align=("relative", 50),
             valign=("relative", 50),
-            width=("relative", 98),
+            width=app.prefs["max_text_width"],
             height=("relative", 60)
         )
 
@@ -1338,6 +1343,9 @@ class MessageBody(urwid.Text):
                     color = str(colornames.index(directive))
                     result.append((color, body))
 
+                elif directive == "dim":
+                    result.append((directive, body))
+
                 elif directive in ["underline", "bold"]:
                     result.append((directive, body))
 
diff --git a/src/formatting.py b/src/formatting.py
index c00b0ec..a3eb73d 100644
--- a/src/formatting.py
+++ b/src/formatting.py
@@ -65,25 +65,28 @@ they are only removed when they occur before a valid expression.
 import re
 
 colors = [
-#0,   1        2        3        4       5        6
-    "red", "yellow", "green", "blue", "cyan", "magenta"
+#0,   1        2        3        4       5        6       dim is not used in color api
+    "red", "yellow", "green", "blue", "cyan", "magenta", "dim"
 ]
 
 markup = [
     "bold", "underline", "linequote", "quote", "rainbow"
 ]
 
-# PS: regex parsing is no longer used for these, preserving anyways
-# tokens being [red: this will be red] and [bold: this will be bold]
-# tokens = re.compile(r"\[(%s): (.+?)]" % "|".join(colors + markup), flags=re.DOTALL)
-# linequotes being chan-style greentext,
-# >like this
-# linequotes = re.compile("^(>.+)$", flags=re.MULTILINE)
 
 # quotes being references to other post_ids, like >>34 or >>0 for OP
 quotes = re.compile(">>([0-9]+)")
-bold = re.compile(r"\*{2}(.{1,20})\*{2}")
-underline = re.compile(r"__(.{1,20})__")
+bold = re.compile(r"(?<!\\)\*{2}(.+?)(?<!\\)\*{2}")
+underline = re.compile(r"(?<!\\)_{2}(.+?)(?<!\\)_{2}")
+escapes = re.compile(r"\\([*_]{2})")
+
+
+def apply_directives(text):
+    # is there a better way to do this? smh....
+    text = quotes.sub(lambda m: "[quote: %s]" % m.group(1), text)
+    text = bold.sub(lambda m: "[bold: %s]" % m.group(1), text)
+    text = underline.sub(lambda m: "[underline: %s]" % m.group(1), text)
+    return escapes.sub(lambda m: m.group(1), text)
 
 
 def parse_segments(text, sanitize_linequotes=True):
@@ -99,19 +102,16 @@ def parse_segments(text, sanitize_linequotes=True):
         for segment in [s.strip() for s in paragraph.split("\n")]:
             if not segment:
                 continue
-            segment = quotes.sub(lambda m: "[quote: %s]" % m.group(1), segment)
-            segment = bold.sub(lambda m: "[bold: %s]" % m.group(1), segment)
-            segment = underline.sub(lambda m: "[underline: %s]" % m.group(1), segment)
-            if segment.startswith(">"):
+            _fp = segment.find(" ")
+            first_word =  segment[:_fp] if _fp != -1 else segment
+            if segment.startswith(">") and not quotes.search(first_word):
                 if sanitize_linequotes:
                     inner = segment.replace("]", "\\]")
                 else:
-                    inner = segment
-                segment = "[linequote: %s]" % inner
-                # pg = pg[0:-1]
-                pg += segment
+                    inner = apply_directives(segment)
+                pg += "[linequote: %s]" % inner
             else:
-                pg += segment + " "
+                pg += apply_directives(segment) + " "
         result.append(pg.strip())
     return result