Implement document matcher parser

3 files changed, 139 insertions, 0 deletions

diff --git a/mammoth/styles/parser/document_matcher_parser.py b/mammoth/styles/parser/document_matcher_parser.py
new file mode 100644
index 0000000..77eae3e
--- /dev/null
+++ b/mammoth/styles/parser/document_matcher_parser.py
@@ -0,0 +1,71 @@
+from ... import documents, document_matchers
+from .tokeniser import TokenType
+from .token_parser import try_parse_class_name, parse_string
+
+
+def parse_document_matcher(tokens):
+    if tokens.try_skip(TokenType.IDENTIFIER, "p"):
+        style_id = try_parse_class_name(tokens)
+        style_name = _parse_style_name(tokens)
+        numbering = _parse_numbering(tokens)
+        
+        return document_matchers.paragraph(
+            style_id=style_id,
+            style_name=style_name,
+            numbering=numbering,
+        )
+        
+    elif tokens.try_skip(TokenType.IDENTIFIER, "r"):
+        style_id = try_parse_class_name(tokens)
+        style_name = _parse_style_name(tokens)
+        
+        return document_matchers.run(
+            style_id=style_id,
+            style_name=style_name,
+        )
+    
+    elif tokens.try_skip(TokenType.IDENTIFIER, "b"):
+        return document_matchers.bold
+    
+    elif tokens.try_skip(TokenType.IDENTIFIER, "i"):
+        return document_matchers.italic
+    
+    elif tokens.try_skip(TokenType.IDENTIFIER, "u"):
+        return document_matchers.underline
+    
+    elif tokens.try_skip(TokenType.IDENTIFIER, "strike"):
+        return document_matchers.strikethrough
+    
+    elif tokens.try_skip(TokenType.IDENTIFIER, "comment-reference"):
+        return document_matchers.comment_reference
+
+
+def _parse_style_name(tokens):
+    if tokens.try_skip(TokenType.SYMBOL, "["):
+        tokens.skip(TokenType.IDENTIFIER, "style-name")
+        tokens.skip(TokenType.SYMBOL, "=")
+        name = parse_string(tokens)
+        tokens.skip(TokenType.SYMBOL, "]")
+        return name
+    else:
+        return None
+
+
+def _parse_numbering(tokens):
+    if tokens.try_skip(TokenType.SYMBOL, ":"):
+        is_ordered = _parse_list_type(tokens)
+        tokens.skip(TokenType.SYMBOL, "(")
+        level = int(tokens.next_value(TokenType.INTEGER)) - 1
+        tokens.skip(TokenType.SYMBOL, ")")
+        return documents.numbering_level(level, is_ordered=is_ordered)
+
+
+def _parse_list_type(tokens):
+    list_type = tokens.next_value(TokenType.IDENTIFIER)
+    if list_type == "ordered-list":
+        return True
+    elif list_type == "unordered-list":
+        return False
+    else:
+        pass
+        # TODO: raise
diff --git a/mammoth/styles/parser/token_iterator.py b/mammoth/styles/parser/token_iterator.py
new file mode 100644
index 0000000..810e50b
--- /dev/null
+++ b/mammoth/styles/parser/token_iterator.py
@@ -0,0 +1,31 @@
+class TokenIterator(object):
+    def __init__(self, tokens):
+        self._tokens = tokens
+        self._index = 0
+    
+    def next_value(self, token_type):
+        return self._next(token_type).value
+    
+    def _next(self, token_type):
+        token = self._tokens[self._index]
+        if token.type == token_type:
+            self._index += 1
+            return token
+        else:
+            raise self._unexpected_token_type(token_type, token)
+    
+    def skip(self, token_type, token_value):
+        token = self._tokens[self._index]
+        if token.type == token_type and token.value == token_value:
+            self._index += 1
+            return True
+        else:
+            raise self._unexpected_token_type(token_type, token)
+    
+    def try_skip(self, token_type, token_value):
+        token = self._tokens[self._index]
+        if token.type == token_type and token.value == token_value:
+            self._index += 1
+            return True
+        else:
+            return False
diff --git a/mammoth/styles/parser/token_parser.py b/mammoth/styles/parser/token_parser.py
new file mode 100644
index 0000000..495d790
--- /dev/null
+++ b/mammoth/styles/parser/token_parser.py
@@ -0,0 +1,37 @@
+import re
+
+from .tokeniser import TokenType
+
+
+def try_parse_class_name(tokens):
+    if tokens.try_skip(TokenType.SYMBOL, "."):
+        return parse_identifier(tokens)
+    else:
+        return None
+
+
+def parse_identifier(tokens):
+    return decode_escape_sequences(tokens.next_value(TokenType.IDENTIFIER))
+
+
+def parse_string(tokens):
+    return decode_escape_sequences(tokens.next_value(TokenType.STRING)[1:-1])
+
+
+_ESCAPE_SEQUENCE_REGEX = re.compile(r"\\(.)")
+
+
+def decode_escape_sequences(value):
+    return _ESCAPE_SEQUENCE_REGEX.sub(_decode_escape_sequence, value)
+    
+    
+def _decode_escape_sequence(match):
+    code = match.group(1)
+    if code == "n":
+        return "\n"
+    elif code == "r":
+        return "\r"
+    elif code == "t":
+        return "\t"
+    else:
+        return code