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

# In[3]:


import pandas as pd
from textblob import Word


# In[4]:


headers = pd.read_csv("header.csv")
headers['Header']


# In[9]:


citation = [Word("citation").synsets[2], Word("reference").synsets[1], Word("cite").synsets[3]]
run = [Word("run").synsets[9],Word("run").synsets[34],Word("execute").synsets[4]]
install = [Word("installation").synsets[0],Word("install").synsets[0],Word("setup").synsets[1],Word("prepare").synsets[0],Word("preparation").synsets[0],Word("manual").synsets[0],Word("guide").synsets[2],Word("guide").synsets[9]]
download = [Word("download").synsets[0]]
requirement = [Word("requirement").synsets[2],Word("prerequisite").synsets[0],Word("prerequisite").synsets[1],Word("dependency").synsets[0],Word("dependent").synsets[0]]
contact = [Word("contact").synsets[9]]
description = [Word("description").synsets[0],Word("description").synsets[1],Word("introduction").synsets[3],Word("introduction").synsets[6],Word("basics").synsets[0],Word("initiation").synsets[1],Word("start").synsets[0],Word("start").synsets[4],Word("started").synsets[0],Word("started").synsets[1],Word("started").synsets[7],Word("started").synsets[8],Word("overview").synsets[0],Word("summary").synsets[0],Word("summary").synsets[2]]
contributor = [Word("contributor").synsets[0]]
documentation = [Word("documentation").synsets[1]]
license = [Word("license").synsets[3],Word("license").synsets[0]]
usage = [Word("usage").synsets[0],Word("example").synsets[0],Word("example").synsets[5],Word("implement").synsets[1],Word("implementation").synsets[1],Word("demo").synsets[1],Word("tutorial").synsets[0],Word("tutorial").synsets[1]]
update = [Word("updating").synsets[0],Word("updating").synsets[3]]
issues = [Word("issues").synsets[0],Word("errors").synsets[5],Word("problems").synsets[0],Word("problems").synsets[2]]
support = [Word("support").synsets[7],Word("help").synsets[0],Word("help").synsets[9],Word("report").synsets[0],Word("report").synsets[6]]



group = dict()
group.update({"citation":citation})
group.update({"download":download})
group.update({"run":run})
group.update({"installation":install})
group.update({"requirement":requirement})
group.update({"contact":contact})
group.update({"description":description})
group.update({"contributor":contributor})
group.update({"documentation":documentation})
group.update({"license":license})
group.update({"usage":usage})
group.update({"update":update})
group.update({"issues":issues})
group.update({"support":support})


def find_sim(wordlist,wd): #returns the max probability between a word and subgroup
    simvalue = []
    for sense in wordlist:
        if(wd.path_similarity(sense)!=None):
            simvalue.append(wd.path_similarity(sense))
    if(len(simvalue)!=0):
        return max(simvalue)
    else:
        return 0
    

def match_group(word_syn,group,threshold):
    currmax = 0
    maxgroup = ""
    simvalues = dict()
    for sense in word_syn: #for a given sense of a word
        similarities = []
        for key, value in group.items(): #value has all the similar words
            path_sim = find_sim(value,sense)
#             print("Similarity is:",path_sim)
            if(path_sim>threshold): #then append to the list
                if(path_sim>currmax):
                    maxgroup = key
                    currmax = path_sim

    return maxgroup


# In[19]:


datadf = pd.DataFrame({'Header': [], 'Group': []})


matchedgroups = []
for h in headers["Header"]:
    sentence = h.split(" ")[1:]
    for s in sentence:
        synn = Word(s).synsets
        if(len(synn)>0):
            bestgroup = match_group(synn,group,0.6)
            if(bestgroup!=""):
                datadf = datadf.append({'Header' : h, 'Group' : bestgroup}, ignore_index=True)
print(datadf)
datadf.to_csv('header_groups.csv', index=False)


# In[ ]:





# In[ ]: