From 64db61bff14593f139ec45e36bbea54305c653bf Mon Sep 17 00:00:00 2001
From: Matthew Honnibal <honnibal@gmail.com>
Date: Fri, 13 Mar 2015 19:21:16 -0400
Subject: [PATCH] * Add Span class to Python API

---
 spacy/tokens.pxd |  6 ++++++
 spacy/tokens.pyx | 33 ++++++++++++++++++++++-----------
 2 files changed, 28 insertions(+), 11 deletions(-)

diff --git a/spacy/tokens.pxd b/spacy/tokens.pxd
index 89902d36d..efc03c368 100644
--- a/spacy/tokens.pxd
+++ b/spacy/tokens.pxd
@@ -54,6 +54,12 @@ cdef class Tokens:
     cdef int set_parse(self, const TokenC* parsed, dict label_ids) except -1
 
 
+cdef class Span:
+    cdef Tokens _seq
+    cdef public int start
+    cdef public int end
+
+
 cdef class Token:
     cdef Vocab vocab
     cdef unicode _string
diff --git a/spacy/tokens.pyx b/spacy/tokens.pyx
index 2ba855554..677f7a101 100644
--- a/spacy/tokens.pyx
+++ b/spacy/tokens.pyx
@@ -237,24 +237,15 @@ cdef class Tokens:
         """This is really only a place-holder for a proper solution."""
         cdef int i
         cdef Tokens sent = Tokens(self.vocab, self._string[self.data[0].idx:])
-        #cdef attr_t period = self.vocab.strings['.']
-        #cdef attr_t question = self.vocab.strings['?']
-        #cdef attr_t exclamation = self.vocab.strings['!']
-        spans = []
         start = None
         for i in range(self.length):
             if start is None:
                 start = i
             if self.data[i].sent_end:
-                spans.append((start, i+1))
+                yield Span(self, start, i+1)
                 start = None
-            #if self.data[i].lex.orth == period or self.data[i].lex.orth == exclamation or \
-            #  self.data[i].lex.orth == question:
-            #    spans.append((start, i+1))
-            #    start = None
         if start is not None:
-            spans.append((start, self.length))
-        return spans
+            yield Span(self, start, self.length) 
 
     cdef int set_parse(self, const TokenC* parsed, dict label_ids) except -1:
         self._py_tokens = [None] * self.length
@@ -267,6 +258,26 @@ cdef class Tokens:
         self._dep_strings = tuple(dep_strings)
 
 
+cdef class Span:
+    """A slice from a Tokens object."""
+    def __cinit__(self, Tokens tokens, int start, int end):
+        self._seq = tokens
+        self.start = start
+        self.end = end
+
+    def __len__(self):
+        if self.end < self.start:
+            return 0
+        return self.end - self.start
+
+    def __getitem__(self, int i):
+        return self._seq[self.start + i]
+
+    def __iter__(self):
+        for i in range(self.start, self.end):
+            yield self._seq[i]
+
+
 cdef class Token:
     """An individual token --- i.e. a word, a punctuation symbol, etc.  Created
     via Tokens.__getitem__ and Tokens.__iter__.