#!/usr/bin/env python
r'''
NAME
    pdf-extract-tables file.pdf > out.txt

SYNOPSIS
    pdf-extract-tables file.pdf > out.txt
      default output format:

    Options:
      -t     cells tagged by column number (_5: value\n) 
                 col sep: "\n"; 
                 white values cells removed
                 row sep: emptyline ("\n\n")
      -tn    col sep: emptyline (\n\n); 
             row sep: double emptyline (\n\n\n);
      -tn1   No tags (no "_5:", just the value)
      -q     quiet: cell 
      -s     skip tables that appear reversed strings

Description

Gets

'''

from jjcli import * 
import pdfplumber

cl=clfilter(opt="1sqtnp:", doc=__doc__)     ## option values in cl.opt dictionary
# FIXME page ranges....  page = int(cl.opt.get("-p",7))

table_settings = {
#      "vertical_strategy": "lines",
#      "horizontal_strategy": "text",
#      "snap_y_tolerance": 5,
#      "intersection_x_tolerance": 15,
}

def guess_rev(tab):
    """ Try to find string reversed evidences (skip table?)"""
    newtab=[]
    for row in tab:
        r = []
        for col in row:
            r.append( col.strip() if col else "")
        newtab.append(r)
    ex1 = [" ".join(x) for x in newtab] 
    ex = " ".join( ex1 )
    notok = len(findall(r' [,.!?;:][a-z]|[a-z][A-Z] |[a-z][cdfpqt] ', ex))
    ok = len(findall(r'[a-z][,.!?;:] | [A-Z][a-z]| [cdfpqt][a-z]', ex))
    ans = notok > ok + 3
    if ans and "-s" in cl.opt:
        return [] 
    elif ans and "-s" in cl.opt:
        tab=[]
        for row in newtab:
            r = []
            for col in row:
                r.append( col[::-1].strip() if col else "")
            tab.append(r)
        return tab
    else:
        return tab


def row2txt(row):
    if "-t" in cl.opt:
        r=""
        for i, x in enumerate(row, start=1):
            if "-1" in cl.opt:
                tag = ""
            else:
                tag = f"_{i}: "
            if not x : continue
            v = sub(r'\n', r'• ', str(x))
            if v.isspace(): continue
            if "-n" in cl.opt: r += f"\n\n{tag}{v}"
            else:              r += f"\n{tag}{v}"
        return r 
    else:
        return " | ".join( map( lambda x: sub(r'\n', r'•',  "" if x is None else str(x)),row))

for filename in cl.args:  
    pdf = pdfplumber.open(filename)

    print( len( pdf.pages)) # go to the required page
    for i, p0 in enumerate(pdf.pages, start=1):
#    p0 = pdf.pages[page] # go to the required page

        tables = p0.extract_tables(table_settings)
        for j,table in enumerate(tables, start=1):
            print(f"\n#------------------------------pg-{i}, table-{j}-")
            tab=guess_rev(table)
            for row in tab:
                v1 = row2txt(row)
                if "-t" in cl.opt:
                    v1 = sub(r'-• *', '', v1)
                    v1 = sub(r'[• ]+', ' ', v1)
                    print(v1)
                elif "-q" in cl.opt:
                    v1 = sub(r'-• *', '', v1)
                    v1 = sub(r'[• ]+', ' ', v1)
                    print(f"\n{v1}")
                else:
                    print(f"____\n{v1}")

'''
#    tables = p0.debug_tablefinder() # list of tables which pdfplumber identifies
#    req_table = tables.tables[0] # Suppose you want to use ith table

    cells = req_table.cells # gives list of all cells in that table
    print(cells)

    for i, cell in enumerate(cells): ## [i:j]: # iterating through the required cells
        print(i, p0.crop(cell).extract_words()) # extract the words 

'''