From 6e99e7b3467064ce08f86c50a458596a4d1bf9c2 Mon Sep 17 00:00:00 2001
From: "James R. Barlow" <jim@purplerock.ca>
Date: Tue, 1 Oct 2019 00:42:58 -0700
Subject: [PATCH] Use lstm_use_matrix for --user-words,patterns

---
 src/ocrmypdf/exec/tesseract.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/src/ocrmypdf/exec/tesseract.py b/src/ocrmypdf/exec/tesseract.py
index 05e3266e..8f88b563 100644
--- a/src/ocrmypdf/exec/tesseract.py
+++ b/src/ocrmypdf/exec/tesseract.py
@@ -269,6 +269,9 @@ def generate_hocr(
     if user_patterns:
         args_tesseract.extend(['--user-patterns', user_patterns])
 
+    if user_words or user_patterns:
+        args_tesseract.extend(['-c', 'lstm_use_matrix=1'])
+
     # Reminder: test suite tesseract spoofers will break after any changes
     # to the number of order parameters here
     args_tesseract.extend([input_file, prefix, 'hocr', 'txt'] + tessconfig)