From e57d55a11d3df86c8da12a27ba6146aa052ef359 Mon Sep 17 00:00:00 2001
From: Tony Cebzanov <[email protected]>
Date: Sat, 13 Jun 2020 20:45:05 -0400
Subject: [PATCH] Add script for detecting character widths and dumping width
 tables

New char_width_tools.py script.  Usage:

    To detect character widths for the current terminal:

        ./char_width_tools.py analyze terminal.csv

    To print the widths of each character in a string:

        ./char_width_tools.py test "string"

    To load a file saved with "analyze" as C/Python source to be pasted
    into str_util.c and old_str_util.python:

        ./char_width_tools.py dump terminal.csv c

        ./char_width_tools.py dump terminal.csv python
---
 bin/char_width_tools.py | 172 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 172 insertions(+)
 create mode 100644 bin/char_width_tools.py

diff --git a/bin/char_width_tools.py b/bin/char_width_tools.py
new file mode 100644
index 0000000..8ab40ed
--- /dev/null
+++ b/bin/char_width_tools.py
@@ -0,0 +1,172 @@
+#!/usr/bin/env python
+
+# Build a table of unicode character widths as rendered by the current terminal.
+
+import sys
+import tty
+import termios
+from select import select
+import re
+import argparse
+import csv
+
+MIN_CODE_POINT = 0
+MAX_CODE_POINT = 0x110000
+
+ESC = "\x1b"
+CLEAR = ESC + "[2J"
+
+BEGINNING_OF_LINE = ESC +"[1;2H"
+
+def getpos():
+
+    buf = ""
+    stdin = sys.stdin.fileno()
+    tattr = termios.tcgetattr(stdin)
+
+    try:
+        tty.setcbreak(stdin, termios.TCSANOW)
+        sys.stdout.write("\x1b[6n")
+        sys.stdout.flush()
+
+        while True:
+            buf += sys.stdin.read(1)
+            if buf[-1] == "R":
+                break
+
+    finally:
+        termios.tcsetattr(stdin, termios.TCSANOW, tattr)
+
+    return [
+        int(x)
+        for x in re.match(r"^\x1b\[(\d*);(\d*)R", buf).groups()
+    ]
+
+
+def get_width(i):
+    c = chr(i)
+    print(BEGINNING_OF_LINE)
+    try:
+        sys.stdout.write(c)
+    except UnicodeEncodeError:
+        return None
+    return getpos()[1]-1
+
+
+def test_string(s):
+
+    widths = []
+    for c in s:
+        widths.append(get_width(ord(c)))
+
+    print(CLEAR)
+
+    for c, w in zip(s, widths):
+        print("%s\t%d" %(c, w))
+    print()
+    print("total\t%d" %(sum(widths)))
+
+def write_table(filename):
+    widths = list()
+    last_width = None
+
+    try:
+        for i in range(MIN_CODE_POINT, MAX_CODE_POINT):
+            w = get_width(i)
+            if w is None:
+                continue
+            if i != MIN_CODE_POINT and w != last_width:
+                widths.append((i-1, last_width))
+            last_width = w
+            if not i % 1000:
+                pct = i / MAX_CODE_POINT
+                print()
+                print("%d/%d (%.02f%%)" %(i, MAX_CODE_POINT, pct*100))
+        widths.append((i, last_width))
+
+    finally:
+        with open(filename, "w") as f:
+            writer = csv.writer(f)
+            for w in widths:
+                writer.writerow("%d\t%d\n" %(w[0], w[1]))
+
+
+def dump(infile, lang):
+
+    widths = list()
+    last_width = None
+
+    with open(infile) as f:
+        reader = csv.reader(f)
+        for row in reader:
+            widths.append((row[0], row[1]))
+
+    if args.lang == "c":
+        print("static const int widths[] = {")
+    elif args.lang == "python":
+        print("widths = [")
+
+    for i, w in widths:
+        if lang == "c":
+            print("    %s, %s," %(i, w))
+
+        elif lang == "python":
+            print("    (%s, %s)," %(i, w))
+        else:
+            raise RuntimeError("unknown language")
+
+    if lang == "c":
+        print("    NULL")
+        print("};")
+    elif lang == "python":
+        print("]")
+
+def main():
+
+    global args
+    parser = argparse.ArgumentParser()
+
+    subparsers = parser.add_subparsers(dest="command", required=True)
+
+    analyze_parser = subparsers.add_parser(
+        "analyze",
+        help="analyze character widths and save them to csv file"
+    )
+    analyze_parser.add_argument(
+        "outfile", metavar="FILE",
+        help="write character widths to this file"
+    )
+    analyze_parser.set_defaults(func=lambda args: write_table(args.outfile))
+
+    test_parser = subparsers.add_parser(
+        "test",
+        help="analyze a set of characters specified on the command line"
+    )
+    test_parser.add_argument(
+        "string", metavar="CHAR",
+        help="character to test"
+    )
+    test_parser.set_defaults(func=lambda args: test_string(args.string))
+
+    dump_parser = subparsers.add_parser(
+        "dump",
+        help="generate source code for width tables from a saved file"
+    )
+    dump_parser.add_argument(
+        "infile", metavar="FILE",
+        help="read character widths from this file"
+    )
+    dump_parser.add_argument(
+        "lang", metavar="LANGUAGE", choices=["c", "python"],
+        help="Language to dump data for."
+    )
+    dump_parser.set_defaults(func=lambda args: dump(args.infile, args.lang))
+
+    args = parser.parse_args()
+
+    print(CLEAR)
+
+    args.func(args)
+
+if __name__ == "__main__":
+    main()
-- 
2.39.0.windows.2