Add support for caclulating confidence measure in libtesseract (Cunei…

…form will default to confidence=0)

src/pyocr/builders.py

@@ -42,7 +42,7 @@ class Box(object):
     was used.
     """
 
-    def __init__(self, content, position, confidence=None):
+    def __init__(self, content, position, confidence=0):
         """
         Arguments:
             content --- a single string
@@ -61,9 +61,8 @@ def get_unicode_string(self):
         This string can be stored in a file as-is (see write_box_file())
         and reread using read_box_file().
         """
-        return to_unicode("%s %s %d %d %d %d") % (
+        return to_unicode("%s %d %d %d %d") % (
             self.content,
-            self.confidence,
             self.position[0][0],
             self.position[0][1],
             self.position[1][0],

src/pyocr/libtesseract/__init__.py

@@ -75,10 +75,11 @@ def get_available_builders():
     ]
 
 
-def _tess_box_to_pyocr_box(box):
+def _tess_box_to_pyocr_box(box, confidence=None):
     return (
         (box[0], box[1]),
         (box[2], box[3]),
+        confidence
     )
 
 
@@ -140,13 +141,17 @@ def image_to_string(image, lang=None, builder=None):
                 res_iterator, lvl_word
             )
 
-            if word is not None and word != "":
+            confidence = tesseract_raw.result_iterator_get_confidence(
+                res_iterator, lvl_word
+            )
+
+            if word is not None and confidence is not None and word != "":
                 (r, box) = tesseract_raw.page_iterator_bounding_box(
                     page_iterator, lvl_word
                 )
                 assert(r)
                 box = _tess_box_to_pyocr_box(box)
-                builder.add_word(word, box)
+                builder.add_word(word, box, confidence)
 
                 if last_word_in_line:
                     builder.end_line()

src/pyocr/libtesseract/tesseract_raw.py

@@ -298,6 +298,12 @@ class OSResults(ctypes.Structure):
     g_libtesseract.TessResultIteratorGetUTF8Text.restype = \
         ctypes.c_void_p
 
+    g_libtesseract.TessResultIteratorConfidence.argtypes = [
+        ctypes.c_void_p,
+        ctypes.c_int,
+    ]
+    g_libtesseract.TessResultIteratorConfidence.restype = ctypes.c_float
+
     g_libtesseract.TessDeleteText.argtypes = [
         ctypes.c_void_p
     ]
@@ -586,6 +592,14 @@ def result_iterator_get_utf8_text(iterator, level):
     g_libtesseract.TessDeleteText(ptr)
     return val
 
+def result_iterator_get_confidence(iterator, level):
+    ptr = g_libtesseract.TessResultIteratorConfidence(
+        ctypes.c_void_p(iterator), level
+    )
+    if ptr is None:
+        return None
+    val = ctypes.c_float(ptr).value
+    return val
 
 def detect_os(handle):
     global g_libtesseract