Skip to content

Python module/script to map Nepali ASCII font faces like Preeti, Sagarmatha, and more to devanagari unicode with direct mapping support for docx and txt files.

License

Notifications You must be signed in to change notification settings

casualsnek/npttf2utf

Repository files navigation

npTTF2UTF

Visits Badge

Python module/script to map Nepali ASCII font faces like Preeti, Sagarmatha, and more to devanagari unicode

This is mainly a python module to help in mapping of various nepali ASCII font faces to its unicode counterpart and unicode or any unicode mappable font back to Preeti font face. It currently supports directly mapping passed strings (fontmapper.py) as well font auto detection and selecting components to map for mapping for docx files (docxhandler.py) and plain text files (txthandler.py).

Requirements

  • python3

Installation

1) Using pip - Pre packaged

$ pip install npttf2utf
        or
$ pip3 install npttf2utf

Or you can package it locally and install. Useful if you want to change something before installing

2) Using pip - Packaged manually

$ pip install build
    or
$ pip3 install build
$ git clone https://github.com/casualsnek/npttf2utf.git
$ cd npttf2utf
$ python3 -m build

This will create a .whl file for installation in 'dist' directory. To install

$ cd dist
$ pip install *.whl
    or
$ pip3 install *.whl

Usage

**0) As WebApp based on npTTF2UTF ( if you just need the work done )

npfontunify is a web app that uses this library to map documents and text. If you just want to convert documents without setting up everything up, there is a demo page in the readme of npfontunify that you can use without setting up or downloading anything !

1) As CLI application

$ npttf2utf -m [mode] -if [origin font] -of [output font] -dc [docx components] -kf [extra known unicode fonts] -i [input file/string] -o [output file] -mf [mapping defination]

Parameters (Same for both) :

Parameter Help/Usage
-h* Shows help and information about the program
-v* Shows version information
-m Usage mode. "string" to pass input string and output on console window, "docx" for working on docx files and "plain" for working with plaintext files
-if The font face which was used for the string or creating the file. In "docx" mode you can use "auto" to autodetect used fonts and map them
-of* The font to which the string or file will be mapped to. Currently, supports 'unicode' and 'Preeti' . Defaults to 'Preeti' if unspecified
-dc* The components of docx which will be processed during mapping. Components are separated by a comma ',' (Defaults to all supported components 'body_paragraph,table,shape')
-kf* While converting docx to Preeti known nepali unicode will be mapped to preeti and other fonts will be ignored. Specify fonts seperated by comma to add more fonts to default list "Kalimati,Mangal,Noto Sans Devanagari"'
-i Input string or path to input file
-o* Path to output file. Not required for "string" mode
-mf* Path to mapping definition file. If not passed it will look for "map.json" in current script directory

Note: The parameters marked with * are optional

Example usage:

  1. To pass string in Preeti in terminal and map it to unicode ("string" mode)
$ npttf2utf -m string -if Preeti -of unicode -i "asdfghjk"

It will map "asdfghjk" to unicode following mapping for Preeti and output "बकमानजवप"

.

  1. To pass string in unicode in terminal and map it to preeti ("string" mode)
$ npttf2utf -m string -if unicode -of Preeti -i "सबिन आचार्य"

It will map "सबिन आचार्य" to Preeti and output ";lag cfrf/\o" (You can also convert other non unicode fonts like Kantipur, Sagarmatha etc to Preeti)

.

  1. To convert docx or txt file ("plain"/"docx" mode) to unicode
$ npttf2utf -m docx -if auto -of unicode -i "document_with_ASCII_font_faces.docx" -o "document_mapped_to_unicode.docx"

It will map the content of document to unicode and save it as "document_mapped_to_unicode.docx" ("auto" as font is available for "docx" mode only)


  1. To convert docx or txt file ("plain"/"docx" mode) to Preeti
$ npttf2utf -m docx -if auto -of Preeti -i "document_with_ASCII_font_faces_or_unicode.docx" -o "document_mapped_to_preeti.docx" -dc "body_paragraph,table,shape" -kf "some,extra,unicode,fonts"

It will map the content of document to unicode and save it as "document_mapped_to_unicode.docx" ("auto" as font is available for "docx" mode only)


2) As python module

$ python
>> import npttf2utf

Class: npttf2utf.FontMapper

"npttf2utf.FontMapper" class can be used to map the fonts to their unicode counterpart. It is also the base for other document converters

Method: __init __

This method initializes the FontMapper class

def __init__(self, map_json):

Returns: None

Argument Description Optional
map_json Path to mapping definition file (Must be readable by current user) False

Method: map_to_unicode

This method maps the passed string to of defined origin font to unicode using the mapping definition

def map_to_unicode(self, string, from_font="Preeti", unescape_html=False):

Returns: String

Argument Description Optional
string String to map False
from_font The origin font in which string was written. Defaults to "Preeti" if not passed True
unescape_html_input Unescape HTML entities in input string before mapping them (Bool) (Defaults to False) True
escape_html_output Escape HTML entities in mapped string before returning them (Bool) (Defaults to False) True

Example usage:

>> import npttf2utf
>> mapper = npttf2utf.FontMapper("npttf2utf/map.json")
>> mapper.map_to_unicode("asdfghjk", from_font="Preeti", unescape_html_input=False, escape_html_output=False)
बकमानजवप
>> 

Method: map_to_preeti

This method maps the passed Devanagari unicode string to preeti using the preetimapper.py

def map_to_preeti(self, string, from_font="Preeti", unescape_html=False):

Returns: String

Argument Description Optional
string String to map False
from_font The origin font in which string was written. Defaults to "Preeti" if not passed True
unescape_html_input Unescape HTML entities in input string before mapping them (Bool) (Defaults to False) True
escape_html_output Escape HTML entities in mapped string before returning them (Bool) (Defaults to False) True

Example usage:

>> import npttf2utf
>> mapper = npttf2utf.FontMapper("npttf2utf/map.json")
>> mapper.map_to_preeti("सबिन आचार्य", from_font="unicode", unescape_html_input=False, escape_html_output=False)
;lag cfrf/\o
>> 

Class: npttf2utf.DocxHandler

"npttf2utf.DocxHandler" class can be used to map docx files to unicode and save them

Method: __init __

This method initializes the DocxHandler class which can be used to map docx files

def __init__(self, rules_file, default_unicode_font_name="Kalimati"):

Returns: None

Argument Description Optional
rules_file Path to mapping definition file (Must be readable by current user) False
default_unicode_font_name The name of font which will be set for a converted segment of docx files. (Defaults to "Kalimati") True

Method: detect_used_fonts

This method returns list of fonts supported by mapping definition which are used in the docx file

def detect_used_fonts(self, docx_file_path):

Returns: List

Argument Description Optional
docx_file_path Path to docx file whose fonts are to be detected False

Method: map_fonts

This method maps the font in docx file and creates new docx file with mapping applied

def map_fonts(self, orginal_file_path, output_file_path="mapped.docx", from_font="auto", to_font="unicode", components=["body_paragraph", "table", "shape"], known_unicode_fonts=[]):

Returns: None

Argument Description Optional
original_file_path Path to docx file whose fonts are to be mapped False
output_file_path Path where the mapped docx file is to saved (Defaults to "mapped.docx") True
from_font The origin font in which string was written. (Defaults to "auto"). "auto" can be passed to detect used font automatically and map them accordingly and leave english characters untouched True
to_font Target for font conversion. (Defaults to "unicode"). Only "unicode" is supported as of now True
components [List] List of components of docx file which will be looked up for text contents. (Defaults to: ["body_paragraph", "table", "shape"]). "body_paragraph", "table" and "shape" are supported as of now True
known_unicode_fonts [List] List of extra nepali unicode font that when detected will be mapped (Only used while mapping to Preeti) True

Example usage:

>> import npttf2utf
>> converter = npttf2utf.DocxHandler("npttf2utf/map.json", default_unicode_font_name="Kalimati")
>> converter.detect_used_fonts("document_with_ASCII_font_faces.docx")
["Preeti", "Sagarmaths"]
>> converter.map_fonts("document_with_ASCII_font_faces.docx", output_file_path="mapped_document.docx", from_font="auto", to_font="unicode", components=["body_paragraph", "table"], known_unicode_fonts=["some", "extra", "nepali", "font"])
>>

Class: npttf2utf.TxtHandler

"npttf2utf.TxtHandler" class can be used to map plain text files to unicode and save them

Method: __init __

This method initializes the TxtHandler class which can be used to map txt files

def __init__(self, rules_file):

Returns: None

Argument Description Optional
rules_file Path to mapping definition file (Must be readable by current user) False

Method: map_fonts

This method maps the font in txt file and creates new txt file with mapping applied

def map_fonts(self, orginal_file_path, output_file_path="mapped.txt", from_font="Preeti", to_font="unicode", components=[], known_unicode_fonts=[]):

Returns: None

Argument Description Optional
original_file_path Path to txt file whose fonts are to be mapped False
output_file_path Path where the mapped txt file is to saved (Defaults to "mapped.txt") True
from_font The origin font in which string was written. (Defaults to "Preeti"). True
to_font Target for font conversion. (Defaults to "unicode"). Only "unicode" is supported as of now True
components Serves no purpose, just there to match the method call of DocxHandler True
known_unicode_fonts Serves no purpose, just there to match the method call of DocxHandler True
>> import npttf2utf
>> converter = npttf2utf.TxTHandler("npttf2utf/map.json")
>> converter.map_fonts("txt_with_ASCII_font_faces.txt", output_file_path="mapped_txt.txt", from_font="Preeti", to_font="unicode", components=[], known_unicode_fonts=[])
>>

Supported docx components

  • Text content in Text boxes/Shapes
  • General paragraphs
  • Text content in table

Supported ASCII font faces

  • Preeti
  • Sagarmatha
  • Kantipur
  • FONTASY_HIMALI_TT
  • PCS NEPALI

Supported Output fonts

  • Devanagari Unicode
  • Preeti

Todos

  • Add support for headers/footers
  • Optimize the code
  • Add support for .odt files
  • Ability to unify fonts to Preeti as well

Adding support for new file type

Create a new file to handle the file type. (You can use docxhandler.py and modify it as needed). The class constructor should take map definition as the first parameter and the file handler class should contain "map_fonts" methods that take original user file, the path for converted file, from the font, to font, and a list of components as arguments. You can map a string to Unicode by using FontMapper.map_to_unicode if the mapping for origin font exists in the definition. (map_to_unicode takes unescape_html argument to that can be used to unescape HTML string before processing and escape it before returning)

Adding mapping for a new font

Open "src/npttf2utf/map.json" and add a JSON key with this structure

"font_name":{
  "version": "v1",
  "rules": {
    "character-map": {
      "character-in-origin-font": "unicode-equivalent-character"
    },
    "pre-rules": [
      ["regex-string-search", "regex-string-replace"]
    ],
    "post-rules": [
      ["regex-string-search", "regex-string-replace"]
    ]
  }
}

pre-rules - Regex find and replace to apply before mapping characters to Unicode character map - Directly mappable character from source font to Unicode. For Preeti: a <-> ब post-rules - The words may not be as expected directly after mapping. So this contains regexes to find them and replace them with corrections (Regex find and replace to apply before mapping characters to Unicode)

Feel free to use this project for any purpose and long as you comply with the license. Any contribution to the project is highly appreciated. If you find any bugs please report it

About

Python module/script to map Nepali ASCII font faces like Preeti, Sagarmatha, and more to devanagari unicode with direct mapping support for docx and txt files.

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages