From 71c57623a286d566266554464f4ad0d9eab238bd Mon Sep 17 00:00:00 2001
From: Dhanush Varma <your@email.com>
Date: Sun, 5 Apr 2026 03:12:38 +0530
Subject: [PATCH] fix: resolve TESSDATA_PREFIX path correctly for all Tesseract
 versions

Two bugs in init_ocr() in ocr.c:

1. The Tesseract 4/5 branch always blindly appended '/tessdata' to the
   path returned by probe_tessdata_location(). If TESSDATA_PREFIX was
   already set to a path ending in 'tessdata/', this caused a double-
   append e.g. '/usr/share/tessdata/tessdata'.

2. The legacy Tesseract <4 branch passed tessdata_path raw to
   TessBaseAPIInit4 without appending 'tessdata' at all, causing
   Tesseract to look for eng.traineddata directly in e.g. '/usr/share/'
   instead of '/usr/share/tessdata/'.

Fix: normalize the path once before both branches. Detect whether the
returned path already ends with 'tessdata' or 'tessdata/', handle
Windows backslash separators, and use the resolved path in both
Tesseract version branches. Add mprint diagnostic for the resolved path.

Fixes #1492
---
 src/lib_ccx/ocr.c | 32 +++++++++++++++++++++++++++++---
 1 file changed, 29 insertions(+), 3 deletions(-)

diff --git a/src/lib_ccx/ocr.c b/src/lib_ccx/ocr.c
index 70fb9227d..9e51601e9 100644
--- a/src/lib_ccx/ocr.c
+++ b/src/lib_ccx/ocr.c
@@ -261,20 +261,46 @@ void *init_ocr(int lang_index)
 	}
 
 	ctx->api = TessBaseAPICreate();
+
+	/* Build the correct tessdata path for TessBaseAPIInit4.
+	 * probe_tessdata_location() returns a base dir e.g. "/opt/homebrew/share/".
+	 * TessBaseAPIInit4 expects "<base>/tessdata" as the data path.
+	 * If TESSDATA_PREFIX already points at the tessdata dir itself,
+	 * avoid appending "tessdata" a second time. */
+	char tess_path[1024];
+	size_t tp_len = strlen(tessdata_path);
+	int already_has_tessdata = (tp_len >= 8 &&
+				    (strcmp(tessdata_path + tp_len - 8, "tessdata/") == 0 ||
+				     strcmp(tessdata_path + tp_len - 8, "tessdata") == 0));
+	if (already_has_tessdata)
+	{
+		snprintf(tess_path, sizeof(tess_path), "%s", tessdata_path);
+	}
+	else
+	{
+		snprintf(tess_path, sizeof(tess_path), "%s%stessdata",
+			 tessdata_path,
+			 (tessdata_path[tp_len - 1] == '/' || tessdata_path[tp_len - 1] == '\\') ? "" : "/");
+	}
+
+	mprint("CCExtractor: using tessdata path: %s\n", tess_path);
+
 	if (!strncmp("4.", TessVersion(), 2) || !strncmp("5.", TessVersion(), 2))
 	{
-		char tess_path[1024];
-		snprintf(tess_path, 1024, "%s%s%s", tessdata_path, "/", "tessdata");
 		if (ccx_options.ocr_oem < 0)
+		{
 			ccx_options.ocr_oem = 1;
+		}
 		ret = TessBaseAPIInit4(ctx->api, tess_path, lang, ccx_options.ocr_oem, NULL, 0, &pars_vec,
 				       &pars_values, 1, false);
 	}
 	else
 	{
 		if (ccx_options.ocr_oem < 0)
+		{
 			ccx_options.ocr_oem = 0;
-		ret = TessBaseAPIInit4(ctx->api, tessdata_path, lang, ccx_options.ocr_oem, NULL, 0, &pars_vec,
+		}
+		ret = TessBaseAPIInit4(ctx->api, tess_path, lang, ccx_options.ocr_oem, NULL, 0, &pars_vec,
 				       &pars_values, 1, false);
 	}