mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2024-12-11 23:19:04 +08:00
Code Issues Actions 10 Packages Projects Releases Wiki Activity

training: Add new flag --workspace_dir to tesstraining_utils.sh

Browse Source 
By default, that script creates two new temporary directories with random
names in /tmp.

The new command line flag --workspace_dir PATH uses the given path as
a base directory for all temporary files.

That allows better reproducable training results (no random directory
names in log files).

Signed-off-by: Stefan Weil <stweil@ub-backup.bib.uni-mannheim.de>
This commit is contained in:
Stefan Weil 2018-07-26 17:07:53 +02:00 committed by Stefan Weil
parent fbff323d6a
commit b19e69086c
1 changed files with 10 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
src/training/tesstrain_utils.sh
View File

@ -18,8 +18,10 @@
if [ "$(uname)" == "Darwin" ];then
FONTS_DIR="/Library/Fonts/"
FONT_CONFIG_CACHE=$(mktemp -d -t font_tmp.XXXXXXXXXX)
else
FONTS_DIR="/usr/share/fonts/"
FONT_CONFIG_CACHE=$(mktemp -d --tmpdir font_tmp.XXXXXXXXXX)
fi
OUTPUT_DIR="/tmp/tesstrain/tessdata"
OVERWRITE=0
@ -148,6 +150,13 @@ parse_flags() {
--wordlist)
parse_value "WORDLIST_FILE" ${ARGV[$j]}
i=$j ;;
--workspace_dir)
rmdir "$FONT_CONFIG_CACHE"
rmdir "$WORKSPACE_DIR"
parse_value "WORKSPACE_DIR" ${ARGV[$j]}
FONT_CONFIG_CACHE=$WORKSPACE_DIR/fc-cache
mkdir -p $FONT_CONFIG_CACHE
i=$j ;;
*)
err_exit "Unrecognized argument ${ARGV[$i]}" ;;
esac
@ -192,11 +201,7 @@ parse_flags() {
# Function initializes font config with a unique font cache dir.
initialize_fontconfig() {
if [[ "$OSTYPE" == "darwin"* ]]; then
export FONT_CONFIG_CACHE=$(mktemp -d -t font_tmp.XXXXXXXXXX)
else
export FONT_CONFIG_CACHE=$(mktemp -d --tmpdir font_tmp.XXXXXXXXXX)
fi
export FONT_CONFIG_CACHE
local sample_path=${FONT_CONFIG_CACHE}/sample_text.txt
echo "Text" >${sample_path}
run_command text2image --fonts_dir=${FONTS_DIR} \