Process input stream line by line

[sqlconvert.git] / mysql2sql / process_tokens.py

diff --git a/mysql2sql/process_tokens.py b/mysql2sql/process_tokens.py
index ac9930eb81963b1e0c376ff67af035aaaca37679..37752e7c146a8d76ad63f9b7ccbbe395066eec4e 100644 (file)
--- a/mysql2sql/process_tokens.py
+++ b/mysql2sql/process_tokens.py
@@ -1,6 +1,8 @@
 
-from sqlparse.sql import TokenList
-from sqlparse.tokens import Name, Error
+from sqlparse import parse
+from sqlparse.sql import Statement
+from sqlparse.tokens import Name, Error, Punctuation, Comment, Newline, \
+    Whitespace
 
 
 def requote_names(token_list):
@@ -22,3 +24,36 @@ def find_error(token_list):
         if token.ttype is Error:
             return True
     return False
+
+
+class StatementGrouper(object):
+    def __init__(self):
+        self.statements = []
+        self.tokens = []
+        self.lines = []
+
+    def process_line(self, line):
+        lines = self.lines
+        lines.append(line)
+        tokens = parse('\n'.join(lines))[0]
+        self.process_tokens(tokens)
+        self.lines = []
+
+    def process_tokens(self, tokens):
+        for token in tokens:
+            self.tokens.append(token)
+            if (token.ttype == Punctuation) and (token.value == ';'):
+                self.statements.append(Statement(self.tokens))
+                self.tokens = []
+
+    def get_statements(self):
+        for statement in self.statements:
+            yield statement
+        self.statements = []
+
+    def close(self):
+        for token in self.tokens:
+            if (token.ttype not in (Comment.Single, Comment.Multiline,
+                                    Newline, Whitespace)):
+                raise ValueError("Incomplete SQL statement: %s" % self.tokens)
+        return self.tokens