Difference between revisions of "Utility:Accent Removal"

Revision as of 22:00, 16 May 2012

Overview

Replacing accented letters with normal ones in the raws fixes this problem.

Some tile sets use the accented characters for additional graphical symbols. This can make racial language text difficult to read. You can remove the accented characters and symbols from the data files. This works on existing worlds and saved games.

Since the structure of language files might change, it is safest if you remove the problem characters from the files yourself. Here are two methods to do just that. The first (Jackard's) only works on Windows, but is probably the easiest for novice users. The second (frobnic8's) will work anywhere Python does (i.e. just about anywhere), but requires using the command line a little.

Jackard's InfoRapid Script

Download Inforapid Search and Replace.

Save the list below to a text file.

Find the following files in DF\raw\objects:

language_DWARF.txt
language_ELF.txt
language_GOBLIN.txt
language_HUMAN.txt

Select them all, right-click and choose 'Search with InfoRapid' from the menu.

Click the Replace tab that shows up in the lower half of the window.

Select your text file from before in the Replace With field, make sure Replace is set to 'Whole Search Expression' and click Start.

A prompt will appear asking for confirmation. Check the Replace All button and click Yes. When the program stops running you are done.

<Command>
	<Search>„</Search>
	<Replace>a</Replace>
</Command>
<Command>
	<Search> </Search>
	<Replace>a</Replace>
</Command>
<Command>
	<Search>ƒ</Search>
	<Replace>a</Replace>
</Command>
<Command>
	<Search>†</Search>
	<Replace>a</Replace>
</Command>
<Command>
	<Search>…</Search>
	<Replace>a</Replace>
</Command>
<Command>
	<Search>‡</Search>
	<Replace>c</Replace>
</Command>
<Command>
	<Search>‰</Search>
	<Replace>e</Replace>
</Command>
<Command>
	<Search>‚</Search>
	<Replace>e</Replace>
</Command>
<Command>
	<Search>Š</Search>
	<Replace>e</Replace>
</Command>
<Command>
	<Search>ˆ</Search>
	<Replace>e</Replace>
</Command>
<Command>
	<Search>‹</Search>
	<Replace>i</Replace>
</Command>
<Command>
	<Search></Search>
	<Replace>i</Replace>
</Command>
<Command>
	<CaseSensitive>Yes</CaseSensitive>
	<Search>¡</Search>
	<Replace>i</Replace>
</Command>
<Command>
	<Search>Œ</Search>
	<Replace>i</Replace>
</Command>
<Command>
	<Search>¤</Search>
	<Replace>n</Replace>
</Command>
<Command>
	<Search>•</Search>
	<Replace>o</Replace>
</Command>
<Command>
	<Search>”</Search>
	<Replace>o</Replace>
</Command>
<Command>
	<Search>“</Search>
	<Replace>o</Replace>
</Command>
<Command>
	<Search>¢</Search>
	<Replace>o</Replace>
</Command>
<Command>
	<Search>—</Search>
	<Replace>u</Replace>
</Command>
<Command>
	<Search>–</Search>
	<Replace>u</Replace>
</Command>
<Command>
	<Search>£</Search>
	<Replace>u</Replace>
</Command>
<Command>
	<Search>˜</Search>
	<Replace>y</Replace>
</Command>

frobnic8's Modified Python Script

If you have the programming language Python installed on your machine (or don't mind installing it) and aren't scared of a command prompt, here is an alternate method. Python comes pre-installed on Mac OS X and almost all distributions of Linux. (If you are using Windows, the command line instructions shown will need to be modified slightly.)

Ensure you have Python installed. (If you have Python 3.x installed, you will need to remove the unicode functions on line 100 and 104, and change the print statements to functions.)

Copy and paste (this modified version of) "The Unicode Hammer" with the name unicode_hammer.py in the raw/objects sub-directory of your Dwarf Fortress directory. (The Unicode Hammer: Is that a name worthy of Dwarf Fortress, or what?)

#!/usr/bin/env python
"""
latin1_to_ascii -- The UNICODE Hammer -- AKA "The Stupid American"

This takes a UNICODE string and replaces Latin-1 characters with
something equivalent in 7-bit ASCII. This returns a plain ASCII string. 
This function makes a best effort to convert Latin-1 characters into 
ASCII equivalents. It does not just strip out the Latin1 characters.
All characters in the standard 7-bit ASCII range are preserved. 
In the 8th bit range all the Latin-1 accented letters are converted to 
unaccented equivalents. Most symbol characters are converted to 
something meaningful. Anything not converted is deleted.

Background:

One of my clients gets address data from Europe, but most of their systems 
cannot handle Latin-1 characters. With all due respect to the umlaut,
scharfes s, cedilla, and all the other fine accented characters of Europe, 
all I needed to do was to prepare addresses for a shipping system.
After getting headaches trying to deal with this problem using Python's 
built-in UNICODE support I gave up and decided to use some brute force.
This function converts all accented letters to their unaccented equivalents. 
I realize this is dirty, but for my purposes the mail gets delivered.

Noah Spurrier noah at noah.org
License free and public domain
"""

"""This version has had its translation table abused to produce
better results for the language files of the game Dwarf Fortress by
frobnic8.

Original here:
http://code.activestate.com/recipes/251871-latin1-to-ascii-the-unicode-hammer/
"""

def latin1_to_ascii (unicrap):
    """This takes a UNICODE string and replaces Latin-1 characters with
    something equivalent in 7-bit ASCII. It returns a plain ASCII string.
    This function makes a best effort to convert Latin-1 characters into
    ASCII equivalents. It does not just strip out the Latin-1 characters.
    All characters in the standard 7-bit ASCII range are preserved.
    In the 8th bit range all the Latin-1 accented letters are converted
    to unaccented equivalents. Most symbol characters are converted to
    something meaningful. Anything not converted is deleted.
    """
    xlate = {
            0x80: 'E',    # Euro sign
            0x81: 'e',    # Blank
            0x82: 'i',    # Single low 9 quote
            0x83: 'f',    # Latin small letter f with hook
            0x84: 'ii',   # Doubel low 9 quote 
            0x85: 'e',    # Horizontal elipsis
            0x86: 't',    # Dagger
            0x87: 'tt',   # Double dagger
            0x88: 'ea',   # Modified circumflex accent
            0x89: 'oloo', # Per mile sign
            0x8a: 'S',    # Latin capital letter S with caron
            0x8b: '<',    # Single left pointing angle quotation
            0x8c: 'OE',   # Latin capital ligature OE
            0x8d: '-',    # Blank
            0x8e: 'Z',    # Latin capital letter Z with caron
            0x8f: '-',    # Blank
            0x90: '-',    # Blank
            0x91: 'ei',   # Left single quote
            0x92: "ie",   # Right single quote
            0x93: 'ii',   # Left double quote
            0x94: "ee",   # Right double quote
            0x95: 'ao',   # Bullet
            0x96: '-',    # En dash
            0x97: '-',    # Em dash
            0x98: '-',    # Small tilde
            0x99: 'TM',   # Trademark sign
            0x9a: 's',    # Latin small letter s with caron
            0x9b: '>',    # Single right pointing andle quotation
            0x9c: 'oe',   # Latin small ligature oe
            0x9d: '-',    # Blank
            0x9e: 'z',    # Latin small letter z with caron
            0x9f: 'Y',    # Latin capital letter Y with diaeresis
            0xa0: '-',    # Non-breaking space
            0xa1: 'i',    # Inverted exclamation mark
            0xa2: 'c',    # Cent sign
            0xa3: 'E',    # Pound sign
            0xa4: 'o',    # Currency sign
            0xa5: 'Y',    # Yen sign
            0xa6: 'l',    # Pipe, broken vertical bar
            0xa7: 'S',    # Section sign
            0xa8: 'ii',   # Spacing diaeresis
            0xa9: 'c',    # Copyright sign
            0xaa: 'a',    # Feminine ordinal indicator
            0xab: '<<',   # Left double angle quotes
            0xac: 'r',    # Not sign
            0xad: '-',    # Soft hypehen
            0xae: 'R',    # Registered trade mark sign
            0xaf: 'aa',   # Spacing macron
            0xb0: 'o',    # Degree sign
            0xb1: 't',    # Plus or minus sign
            0xb2: '2',    # Superscript 2
            0xb3: '3',    # Superscript 3
            0xb4: "'",    # Acute accent
            0xb5: 'u',    # Micro sign
            0xb6: 'P',    # Pilcrow sign
            0xb7: 'o',    # Middle dot
            0xb8: 'e',    # Cedilla
            0xb9: '1',    # Superscript 1
            0xba: 'o',    # Masculine ordinal indicator
            0xbb: '>>',   # Right double angle quotes
            0xbc: '1/4',  # Fraction one quarter
            0xbd: '1/2',  # Fraction one half
            0xbe: '3/4',  # Fraction three quarters
            0xbf: 'b',    # Inverted question mark
            0xc0: 'A',    # Latin capital letter A with grave
            0xc1: 'A',    # Latin capital letter A with acute
            0xc2: 'A',    # Latin capital letter A with circumflex
            0xc3: 'A',    # Latin capital letter A with tilde
            0xc4: 'A',    # Latin capital letter A with diaeresis
            0xc5: 'A',    # Latin capital letter A with ring above
            0xc6: 'Ae',   # Latin capital letter AE
            0xc7: 'C',    # Latin capital letter C with cedilla
            0xc8: 'E',    # Latin capital letter E with grave
            0xc9: 'E',    # Latin capital letter E with acute
            0xca: 'E',    # Latin capital letter E with circumflex
            0xcb: 'E',    # Latin capital letter E with diaeresis
            0xcc: 'I',    # Latin capital letter I with grave
            0xcd: 'I',    # Latin capital letter I with acute
            0xce: 'I',    # Latin capital letter I with circumflex
            0xcf: 'I',    # Latin capital letter I with diaeresis
            0xd0: 'D',    # Latin capital letter ETH
            0xd1: 'N',    # Latin capital letter N with tilde
            0xd2: 'O',    # Latin capital letter O with grave
            0xd3: 'O',    # Latin capital letter O with acute
            0xd4: 'O',    # Latin capital letter O with circumflex
            0xd5: 'O',    # Latin capital letter O with tilde
            0xd6: 'O',    # Latin capital letter O with diaeresis
            0xd7: 'x',    # Multiplication sign
            0xd8: 'O',    # Latin capital letter O with slash
            0xd9: 'U',    # Latin capital letter U with grave
            0xda: 'U',    # Latin capital letter U with acute
            0xdb: 'U',    # Latin capital letter U with circumflex
            0xdc: 'U',    # Latin capital letter U with diaeresis
            0xdd: 'Y',    # Latin capital letter Y with acute
            0xde: 'P',    # Latin capital letter THORN
            0xdf: 'B',    # Latin small letter sharp s
            0xe0: 'a',    # Latin small letter a with grave
            0xe1: 'a',    # Latin small letter a with acute
            0xe2: 'a',    # Latin small letter a with circumflex
            0xe3: 'a',    # Latin small letter a with tilde
            0xe4: 'a',    # Latin small letter a with diaeresis
            0xe5: 'a',    # Latin small letter a with ring above
            0xe6: 'ae',   # Latin small letter ae
            0xe7: 'c',    # Latin small letter c with cedilla
            0xe8: 'e',    # Latin small letter e with grave
            0xe9: 'e',    # Latin small letter e with acute
            0xea: 'e',    # Latin small letter e with circumflex
            0xeb: 'e',    # Latin small letter e with diaeresis
            0xec: 'i',    # Latin small letter i with grave
            0xed: 'i',    # Latin small letter i with acute
            0xee: 'i',    # Latin small letter i with circumflex
            0xef: 'i',    # Latin small letter i with diaeresis
            0xf0: 'oa',   # Latin small letter eth
            0xf1: 'n',    # Latin small letter n with tilde
            0xf2: 'o',    # Latin small letter o with grave
            0xf3: 'o',    # Latin small letter o with acute
            0xf4: 'o',    # Latin small letter o with circumflex
            0xf5: 'o',    # Latin small letter o with diaeresis
            0xf6: 'o',    # Latin small letter o with slash
            0xf7: 'l',    # Division sign
            0xf8: 'o',    # Latin small letter o with 
            0xf9: 'u',    # Latin small letter u with grave
            0xfa: 'u',    # Latin small letter u with acute
            0xfb: 'u',    # Latin small letter u with circumflex
            0xfc: 'u',    # Latin small letter u with diaeresis
            0xfd: 'y',    # Latin small letter y with acute
            0xfe: 'p',    # Latin small letter thorn
            0xff: 'y',    # Latin small letter y with diaeresis
            }


    r = ''
    for i in unicrap:
        if ord(i) in xlate:
            r += xlate[ord(i)]
        elif ord(i) >= 0x80:
            pass
        else:
            r += str(i)
    return r

if __name__ == '__main__':
    import sys
    input = sys.stdin
    output = sys.stdout
    if len(sys.argv) == 1 or (len(sys.argv) == 2 and \
       sys.argv[1] in ('-h', '-H', '-?', '--help', '/?', '/H', '/h')):
        print 'unicode_hammer.py [infile [outfile]]\n'
        #for python 3.x, changes the following line to s = ''
        s = unicode('','latin-1')
        for c in range(32, 256):
            if c != 0x7f:
                #for python 3.x, change the following line to s += str(chr(c))
                s += unicode(chr(c), 'latin-1')
            plain_ascii = latin1_to_ascii(s)

        #for python 3.x, change all of the following print statements to functions (wrap the entire statement in parenthesis)
        print 'INPUT type:', type(s)
        print 'INPUT:'
        print s.encode('latin-1')
        print
        print 'OUTPUT type:', type(plain_ascii)
        print 'OUTPUT:'
        print plain_ascii
        sys.exit()

    if len(sys.argv) > 1:
        input = open(sys.argv[1])
    if len(sys.argv) > 2:
        output = open(sys.argv[2], 'w')
    for line in input:
        output.write(latin1_to_ascii(line))

Open a command prompt and change directory to your raw/objects directory.

Rename the four language files, adding '.orig' to the end of their names:

mv language_DWARF.txt language_DWARF.txt.orig
mv language_ELF.txt language_ELF.txt.orig
mv language_GOBLIN.txt language_GOBLIN.txt.orig
mv language_HUMAN.txt language_HUMAN.txt.orig

Apply the hammer to each of the four language files as follows:

python unicode_hammer.py language_DWARF.txt.orig language_DWARF.txt
python unicode_hammer.py language_ELF.txt.orig language_ELF.txt
python unicode_hammer.py language_GOBLIN.txt.orig language_GOBLIN.txt
python unicode_hammer.py language_HUMAN.txt.orig language_HUMAN.txt

Enjoy!

The Linux way

Conversion between character sets is a standard part of Linux. To convert all the files in one go, change to the "raw/objects" directory and run this command:

for f in language_*.txt; do \
  iconv -f CP437 -t ASCII//TRANSLIT $f > $f.new; \
  mv -fv $f.new $f; \
done

All accented characters are converted to their normal, non-accented versions. Other characters (if any) are converted to their closest 7-bit ASCII representation.

This will overwrite the original language files. If you want them back, you can always unzip them again:

unzip -j path-to-zipfile raw/objects/language_\*.txt

Hermanos small app

For Windows users there is this small application that replaces accented characters from files by just dragging & dropping the file on the application icon.