#!/usr/bin/env python
# coding: utf-8

# <a href="http://laf-fabric.readthedocs.org/en/latest/" target="_blank"><img align="left" src="images/laf-fabric-xsmall.png"/></a>
# <a href="http://www.godgeleerdheid.vu.nl/etcbc" target="_blank"><img align="left" src="images/VU-ETCBC-xsmall.png"/></a>
# <a href="http://tla.mpi.nl" target="_blank"><img align="right" src="images/TLA-xsmall.png"/></a>
# <a href="http://www.dans.knaw.nl" target="_blank"><img align="right"src="images/DANS-xsmall.png"/></a>

# <img align="left" src="images/1kings-syriac-shot.png"/>

# # CALAP

# **About CALAP**. CALAP stands for Computer-Assisted Linguistic Analysis of the Peshitta.
# CALAP was a [project](https://openaccess.leidenuniv.nl/handle/1887/10866) at the University of Leiden.
# The [Peshitta](http://en.wikipedia.org/wiki/Peshitta) is a collection of Syriac texts. According to Wikipedia it is the standard version of the Bible in churches of the Syriac tradition. Resources can be found on [peshitta.org](http://www.peshitta.org). 
# 
# The text we use below comes from the [Peshitta Institute Leiden](http://www.hum.leiden.edu/religion/research/peshitta-institute/peshitta-institute.html), and has been prepared as an EMDROS database, which is now held by the [ETCBC](http://www.godgeleerdheid.vu.nl/etcbc). 
# 
# From there is has been converted to [LAF](http://www.iso.org/iso/catalogue_detail.htm?csnumber=37326)
# by Dirk Roorda, and this notebook accesses this LAF data by means of
# [LAF-Fabric](http://laf-fabric.readthedocs.org/en/latest/).
# 
# The LAF-data of the CALAP project has been archived at DANS:
# [DOI 10.17026/dans-zv9-w9d2](http://dx.doi.org/10.17026/dans-zv9-w9d2)

# # Text from features

# Here comes the plain text of the CALAP data.
# 
# The CALAP database only contains the surface consonants as textual representation.

# In[1]:


import sys
import collections

from etcbc.lib import Transcription
from laf.fabric import LafFabric
fabric = LafFabric()


# In[2]:


fabric.load('calap', '--', 'plain', {
    "xmlids": {"node": False, "edge": False},
    "features": ('''
        otype
        surface_consonants
        psp
        book chapter verse verse_label
    ''',''),
    "primary": True,
})
exec(fabric.localnames.format(var='fabric'))


# In[3]:


plain_file = outfile("calap_plain.txt")

tr = Transcription()
catalog = set()
for i in F.otype.s('word'):
    sf = F.surface_consonants.v(i)
    for x in sf: catalog.add(x)
    the_text = tr.to_syriac(sf)
    plain_file.write(the_text + ' ')

plain_file.close()
print(','.join(sorted(catalog)))


# This file does not have newlines, it is a blob of consonant transcriptions for each word separated by spaces.

# ## Passage indicators

# If you want books, chapters and verses marked, you can achieve it in the following way:

# In[4]:


plainx_file = outfile("calap_plainx.txt")

the_book = None
the_chapter = None
the_verse = None

for i in NN():
    this_type = F.otype.v(i)
    if this_type == "word":
        the_text = tr.to_syriac(F.surface_consonants.v(i))
        the_suffix = ' '
        plainx_file.write(the_text + the_suffix)
    elif this_type == "book":
        the_book = F.book.v(i)
        sys.stderr.write("\r{:>6} {:<30}".format(i, the_book)) 
        plainx_file.write("\n{}".format(the_book))
    elif this_type == "chapter":
        the_chapter = F.chapter.v(i)
        plainx_file.write("\n{} {}".format(the_book, the_chapter))
    elif this_type == "verse":
        the_verse = F.verse.v(i)
        plainx_file.write("\n{}:{} ".format(the_chapter, the_verse))
sys.stderr.write("\n")

plainx_file.close()


# In order to show the syriac text, you need to install a font that has glyphs for the syriac unicode characters (0700 - 074F).
# For example: Estrangelo Edessa from [Meltho](http://www.bethmardutho.org/index.php/resources/fonts.html).

# In[5]:


get_ipython().system("head -n 10 {my_file('calap_plainx.txt')}")


# If you are in an environment where you do not have this font installed, see the screenshot at the top screenshot.
# 

# ## Verse list

# We can get the text in a quite different way: just read it from the *primary data*.
# 
# Let us do that per verse.

# In[6]:


verse_file = outfile("calap_verses.txt")

for i in F.otype.s('verse'):
    the_text = tr.to_syriac(''.join([txt for (j, txt) in P.data(i)]))
    the_verse = F.verse_label.v(i)
    verse_file.write("{}\n{}\n".format(the_verse, the_text))

verse_file.close()


# In[7]:


get_ipython().system("head -n 10 {my_file('calap_verses.txt')}")


# ## Empty words

# In the BHS there are words that have an empty representation.
# 
# Let us have a closer look to the CALAP.
# Are there empty words?

# In[8]:


ewords = collections.defaultdict(lambda: [])
verse = None

for i in NN(test=F.otype.v, values=['verse', 'word']):
    if F.otype.v(i) == 'verse':
        verse = i
        continue
    text = F.surface_consonants.v(i)
    if text == '':
        lex = lexeme.v(i)
        pos = F.psp.v(i)
        ewords[(lex, pos)].append(verse)

for (item, occs) in sorted(ewords.items(), key=lambda x: (-len(x[1]), x[0][1], x[0][0])):
    print("{:>6} x {:<15} = {:>10} in {}{}".format(
        len(occs), 
        item[1], 
        item[0], 
        "; ".join([F.verse_label.v(j) for j in occs][0:5]),
        ' ...' if len(occs) > 20 else '',
    ))
if not len(ewords):
    print("No empty words found")


# In[9]:


close()


# In[ ]: