From 106b0b9a79175cb60a4fe426a8b779623d24fbe2 Mon Sep 17 00:00:00 2001
From: Matthw Honnibal <honnibal+gh@gmail.com>
Date: Sat, 2 Nov 2019 12:47:10 +0100
Subject: [PATCH] Support GloVe vectors txt format in init-model

---
 spacy/cli/init_model.py | 17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/spacy/cli/init_model.py b/spacy/cli/init_model.py
index c285a12a6..f2114085a 100644
--- a/spacy/cli/init_model.py
+++ b/spacy/cli/init_model.py
@@ -206,7 +206,7 @@ def read_vectors(vectors_loc):
     from tqdm import tqdm
 
     f = open_file(vectors_loc)
-    shape = tuple(int(size) for size in next(f).split())
+    shape, f = _get_shape(f)
     vectors_data = numpy.zeros(shape=shape, dtype="f")
     vectors_keys = []
     for i, line in enumerate(tqdm(f)):
@@ -220,6 +220,21 @@ def read_vectors(vectors_loc):
     return vectors_data, vectors_keys
 
 
+def _get_shape(file_):
+    """Return a tuple with (number of entries, vector dimensions). Handle
+    both word2vec/FastText format, which has a header with this, or GloVe's
+    format, which doesn't."""
+    first_line = next(file_).split()
+    if len(first_line) == 2:
+        return tuple(int(size) for size in first_line), file_
+    count = 1
+    for line in file_:
+        count += 1
+    file_.seek(0)
+    shape = (count, len(first_line)-1)
+    return shape, file_
+
+
 def read_freqs(freqs_loc, max_length=100, min_doc_freq=5, min_freq=50):
     # temp fix to avoid import issues cf https://github.com/explosion/spaCy/issues/4200
     from tqdm import tqdm