Add better support for nested entities (both for HTML and Markdown) (#297)

* Added better support for nested entities, both for HTML and Markdown

* Tiny style fix

* Make use of pre-defined constants

pyrogram/client/parser/html.py

@@ -147,43 +147,38 @@ class HTML:
     @staticmethod
     def unparse(text: str, entities: list):
         text = utils.add_surrogates(text)
-        copy = text
+
+        entities_offsets = []
 
         for entity in entities:
+            entity_type = entity.type
             start = entity.offset
             end = start + entity.length
 
-            type = entity.type
+            if entity_type in ("bold", "italic", "underline", "strike"):
-
+                start_tag = "<{}>".format(entity_type[0])
-            url = entity.url
+                end_tag = "</{}>".format(entity_type[0])
-            user = entity.user
+            elif entity_type in ("code", "pre", "blockquote"):
-
+                start_tag = "<{}>".format(entity_type)
-            sub = copy[start:end]
+                end_tag = "</{}>".format(entity_type)
-
+            elif entity_type == "text_link":
-            if type == "bold":
+                url = entity.url
-                style = "b"
+                start_tag = '<a href="{}">'.format(url)
-            elif type == "italic":
+                end_tag = "</a>"
-                style = "i"
+            elif entity_type == "text_mention":
-            elif type == "underline":
+                user = entity.user
-                style = "u"
+                start_tag = '<a href="tg://user?id={}">'.format(user.id)
-            elif type == "strike":
+                end_tag = "</a>"
-                style = "s"
-            elif type == "code":
-                style = "code"
-            elif type == "pre":
-                style = "pre"
-            elif type == "blockquote":
-                style = "blockquote"
-            elif type == "text_link":
-                text = text[:start] + text[start:].replace(sub, '<a href="{}">{}</a>'.format(url, sub), 1)
-                continue
-            elif type == "text_mention":
-                text = text[:start] + text[start:].replace(
-                    sub, '<a href="tg://user?id={}">{}</a>'.format(user.id, sub), 1)
-                continue
             else:
                 continue
 
-            text = text[:start] + text[start:].replace(sub, "<{0}>{1}</{0}>".format(style, sub), 1)
+            entities_offsets.append((start_tag, start,))
+            entities_offsets.append((end_tag, end,))
+
+        # sorting by offset (desc)
+        entities_offsets.sort(key=lambda x: -x[1])
+
+        for entity, offset in entities_offsets:
+            text = text[:offset] + entity + text[offset:]
 
         return utils.remove_surrogates(text)

pyrogram/client/parser/markdown.py

@@ -107,44 +107,44 @@ class Markdown:
     @staticmethod
     def unparse(text: str, entities: list):
         text = utils.add_surrogates(text)
-        copy = text
+
+        entities_offsets = []
 
         for entity in entities:
+            entity_type = entity.type
             start = entity.offset
             end = start + entity.length
 
-            type = entity.type
+            if entity_type == "bold":
-
+                start_tag = end_tag = BOLD_DELIM
-            url = entity.url
+            elif entity_type == "italic":
-            user = entity.user
+                start_tag = end_tag = ITALIC_DELIM
-
+            elif entity_type == "underline":
-            sub = copy[start:end]
+                start_tag = end_tag = UNDERLINE_DELIM
-
+            elif entity_type == "strike":
-            if type == "bold":
+                start_tag = end_tag = STRIKE_DELIM
-                style = BOLD_DELIM
+            elif entity_type == "code":
-            elif type == "italic":
+                start_tag = end_tag = CODE_DELIM
-                style = ITALIC_DELIM
+            elif entity_type in ("pre", "blockquote"):
-            elif type == "underline":
+                start_tag = end_tag = PRE_DELIM
-                style = UNDERLINE_DELIM
+            elif entity_type == "text_link":
-            elif type == "strike":
+                url = entity.url
-                style = STRIKE_DELIM
+                start_tag = "["
-            elif type == "code":
+                end_tag = "]({})".format(url)
-                style = CODE_DELIM
+            elif entity_type == "text_mention":
-            elif type == "pre":
+                user = entity.user
-                style = PRE_DELIM
+                start_tag = "["
-            # TODO: Blockquote for MD
+                end_tag = "](tg://user?id={})".format(user.id)
-            # elif type == "blockquote":
-            #     style = ...
-            elif type == "text_link":
-                text = text[:start] + text[start:].replace(sub, '[{1}]({0})'.format(url, sub), 1)
-                continue
-            elif type == "text_mention":
-                text = text[:start] + text[start:].replace(
-                    sub, '[{1}](tg://user?id={0})'.format(user.id, sub), 1)
-                continue
             else:
                 continue
 
-            text = text[:start] + text[start:].replace(sub, "{0}{1}{0}".format(style, sub), 1)
+            entities_offsets.append((start_tag, start,))
+            entities_offsets.append((end_tag, end,))
+
+        # sorting by offset (desc)
+        entities_offsets.sort(key=lambda x: -x[1])
+
+        for entity, offset in entities_offsets:
+            text = text[:offset] + entity + text[offset:]
 
         return utils.remove_surrogates(text)