#!/bin/bash
get_line_count() {
wc -l "$1" | cut -d' ' -f1
}
process_file() {
local INPUT_FILE=$1
local FORCE=$2
local OUTPUT_FILE="${INPUT_FILE}.metta"
local HEAD=$(basename -- "$INPUT_FILE")
HEAD=${HEAD/.tsv/}
HEAD=${HEAD/.fb/}
# HEAD=${HEAD%.*} # Remove the file extension
HEAD=${HEAD/_fb_2[0-9][0-9][0-9]_[0-9][0-9]/}
HEAD=${HEAD/2[0-9][0-9][0-9]_[0-9][0-9]/}
HEAD=${HEAD/\./_}
# Skip if the input file is already a .metta file
if [[ "$INPUT_FILE" == *.metta ]]; then
#echo "Skipping .metta file: $INPUT_FILE"
return
fi
local input_lines=$(get_line_count "$INPUT_FILE")
echo -e "Into-MeTTa: $input_lines lines. \t $INPUT_FILE "
echo -en " "
# Check if the output file exists and is newer than the input file
if [[ "$FORCE" != "--force" ]] && [ -e "$OUTPUT_FILE" ] && [ "$OUTPUT_FILE" -nt "$INPUT_FILE" ]; then
local line_count=$(get_line_count "$OUTPUT_FILE")
# Calculate 80% of the input line count
local threshold=$(( input_lines * 80 / 100 ))
if [ "$line_count" -ge "$threshold" ] && [ "$line_count" -ge "400" ]; then
echo -e "$line_count atoms. \t Skipping: $HEAD already exists and is newer."
return
fi
fi
local start_time=$(date +%s.%N) # Get the start time with nanosecond precision
dos2unix -f "$INPUT_FILE" > /dev/null 2>/dev/null
# Process based on the file extension
case "$INPUT_FILE" in
*.obo | *.json | *.fa)
rm -f "$OUTPUT_FILE"
echo -ne "."
###########swipl -l src/main/flybase_convert.pl -- --convert "$INPUT_FILE" --halt | tee -p "$OUTPUT_FILE"
#echo "" ; echo ""
#echo "swipl -l src/main/flybase_convert.pl -- --context=$HEAD --convert \"$INPUT_FILE\" --halt > \"$OUTPUT_FILE\""
#echo "" ; echo ""
swipl -l src/main/flybase_convert.pl -- --context=$HEAD --convert "$INPUT_FILE" --halt #> "$OUTPUT_FILE" 2>/dev/null
echo -ne "."
# Create a temporary file
temp_file=$(mktemp)
echo -ne "."
# Make sure temporary file will be deleted on script exit
trap 'rm -f "$temp_file"' EXIT
# Remove duplicates with awk and save to temporary file
awk '!seen[$0]++' "$OUTPUT_FILE" > "$temp_file"
# Overwrite the original file with the temporary file
mv "$temp_file" "$OUTPUT_FILE"
echo -ne "."
;;
*)
echo -ne "."
#python3 src/mettalog.py "$INPUT_FILE" --analyze > "$OUTPUT_FILE"
python3 src/panda_util.py "$INPUT_FILE" --analyze > "$OUTPUT_FILE"
skip_lines=$?
echo -ne "."
awk -v head="$HEAD" -v skip="$skip_lines" '
function should_quote_metta(field) {
if (field ~ /[ ()\n\t\r]/) {
return 1 # Quote if field contains space, (, ), newline, tab, or carriage return
}
if (field ~ /[^[:print:]]/) {
return 1 # Quote if field contains any non-printing character
}
if (field ~ /[" ]|'\''|\/|,|\|/) {
# return 1 # Quote if field contains ", space, , /, , or |
}
if (field ~ /^[0-9]/ && field !~ /^[0-9]+$/) {
# return 1 # Quote if field starts with a number and contains non-numeric characters
}
return 0 # No need to quote
}
BEGIN {FS="\t"; OFS=" "}
{
# Skip lines that start with whitespace (if any) followed by # or ;
if (/^[[:space:]]*[#;]/) {
next
}
# Skip lines that does not contain a tab character (thus having less than 2 fields)
if (NF < 2) {
next
}
# Construct the line to check for uniqueness
line_output = "(" head
non_blank_found = 0
for (i = 1; i <= NF; i++) {
# Remove all control characters except newline
gsub(/[[:cntrl:]]/, "", $i)
# Escape double quotes to avoid breaking the output syntax
gsub(/"/, "\\\"", $i)
if ($i == "" || $i ~ /^[[:space:]]*$/) {
# Represent fields that are only whitespace as ()
line_output = line_output " ()"
} else {
# Determine if we should quote this field
if (should_quote_metta($i)) {
line_output = line_output " \"" $i "\""
} else {
line_output = line_output " " $i
}
non_blank_found = 1
}
}
# Skip this line if it is a duplicate
if (non_blank_found && !(line_output in printed_lines)) {
if (skip > 0) {
skip--
} else {
print line_output ")"
printed_lines[line_output] = 1
}
}
}
' "$INPUT_FILE" >> "$OUTPUT_FILE"
;;
esac
local end_time=$(date +%s.%N) # Get the end time with nanosecond precision
local elapsed_time=$(echo "$end_time - $start_time" | bc) # Calculate elapsed time using bc
elapsed_time=$(printf "%.2f" "$elapsed_time") # Format the elapsed time
local output_lines=$(get_line_count "$OUTPUT_FILE")
echo -e "$output_lines atoms. \t Converted $HEAD in $elapsed_time seconds."
}
# Check if any arguments are provided
if [ "$#" -eq 0 ]; then
echo "Usage: $0 [--force] <list of directories and files>"
exit 1
fi
FORCE=""
ARGS=()
# Separate the --force flag from the other arguments
for arg in "$@"; do
if [[ "$arg" == "--force" ]]; then
FORCE="--force"
else
ARGS+=("$arg")
fi
done
# Process each argument
for INPUT_PATH in "${ARGS[@]}"; do
if [ -d "$INPUT_PATH" ]; then
# Start timing for directory processing
dir_start_time=$(date +%s.%N)
echo "Processing TSV files: $INPUT_PATH"
find "$INPUT_PATH" -type f -name "*.*" -not -name "*.metta" -not -name "*.obo" -not -name "*.json" -not -name "*.fa" -print0 | while IFS= read -r -d $'\0' file; do
process_file "$file" "$FORCE"
done
echo "Processing Special files: $INPUT_PATH"
find "$INPUT_PATH" -type f \( -name "*.obo" -or -name "*.json" -or -name "*.fa" \) -print0 | while IFS= read -r -d $'\0' file; do
echo process_file "$file" "$FORCE"
done
# End timing for directory processing
dir_end_time=$(date +%s.%N)
dir_elapsed_time=$(echo "$dir_end_time - $dir_start_time" | bc)
dir_elapsed_time=$(printf "%.2f" "$dir_elapsed_time")
echo "Completed processing of directory '$INPUT_PATH' in $dir_elapsed_time seconds."
elif [ -f "$INPUT_PATH" ]; then
# Show elapsed time for individual files
process_file "$INPUT_PATH" "$FORCE"
head -100 "$INPUT_PATH".metta
tail -10 "$INPUT_PATH".metta
else
echo "Error: '$INPUT_PATH' is not a valid file or directory."
fi
done