This script parses text interviews for certain words/phrases. It writes a DataFrame to a csv file containing the word/phrase searched for, subject ID number, subject gender, subject age, subject gender, appearance number, the phrase before and after, and the word before and after.

span_ling.py

# =====================================================
# This script was created to help parse the information
# in a text interview programmatically as opposed to 
# doing so by hand. Currently, this program will 
# find a specifed word, number the first ocurrence of
# the word, and return the phrase before and after the 
# word. 
# =====================================================

import os 
import re
import pandas as pd 
import numpy as np 
current_path = "~/Documents/Transcriptions/Transcriptions/"
dataframe_path = "~/Documents/Transcriptions/DataFrames/"

# this function gets the 
# text file read into Python
# --------------------------
def text_grabber(file_path, file_name):
    file = open(os.path.expanduser(file_path + file_name), encoding = "utf-8")
    lines = file.readlines()
    return(lines)

# this function helps us count the number
# of times that a word appears 
# in the entire text
# ----------------------------
def word_counter(text_object, word_to_find):
    word_count = 0
    for line in text_object:
        line = line.lower()
        for word in line.split():
            if re.findall(word_to_find, word):
                word_count = word_count + 1
    return(word_count)

# This function returns the 
# appearance number in an array.  
# ------------------------------
def word_num(text_object, word_to_find):
    num_array = []
    word_num = 1
    for line in text_object:
        line = line.lower() # convert to lower case
        if re.findall(word_to_find, line):
            num_array.append(word_num)
            word_num = word_num + 1
    return(num_array)

# This function gets the phrase before
# ------------------------------------
def get_before(text_object, word_to_find):
    num_array = []
    before = 0 
    for line in text_object:
        line = line.lower()
        if re.findall(word_to_find, line):
            word_part = line.partition(word_to_find)
            phrase_before = word_part[before]
            num_array.append(phrase_before)
    return(num_array)

# This function gets 
# the phrase after a word 
# ------------------------
def get_after(text_object, word_to_find):
    num_array = []
    after = 2 
    for line in text_object:
        line = line.lower()
        if re.findall(word_to_find, line):
            word_part = line.partition(word_to_find)
            phrase_after = word_part[after]
            num_array.append(phrase_after)
    return(num_array)

# this function gets the 
# immediate word before a word
# ----------------------------
def get_word_before(text_object, word_to_find):
    num_array = []
    before = 0 
    for line in text_object:
        line = line.lower()
        line = line.strip()
        if re.findall(word_to_find, line):
            word_part = line.partition(word_to_find)
            phrase_before = word_part[before]
            phrase_split = phrase_before.split()
            # add exception handling for places where
            # the phrase is on the previous line
            try: 
                word_before = phrase_split[len(phrase_split) - 1]
            except IndexError:
                word_before = "NA"
            num_array.append(word_before)
    return(num_array)

# this function gets the 
# immediate word after a given word 
# ---------------------------------
def get_word_after(text_object, word_to_find):
    num_array = []
    after = 2
    for line in text_object:
        line = line.lower()
        line = line.strip()
        if re.findall(word_to_find, line):
            word_part = line.partition(word_to_find)
            phrase_after = word_part[after]
            phrase_split = phrase_after.split()
            # add exception handling for places where
            # the phrase is on the next line
            try: 
                word_after = phrase_split[0]
            except IndexError:
                word_after = "NA"
            num_array.append(word_after)
    return(num_array)

# this function assigns an ID number for use when 
# later combining dataframe together. 
# ------------------------------------------------
def assign_id_number(text_object, word_to_find, ID_num):
    num_array = []
    i = 0
    while i < len(word_num(text_object, word_to_find)):
        num_array.append(ID_num)
        i = i + 1 
    return(num_array)

# this function lists the word that we are looking at
# so there is more context 
# ---------------------------------------------------
def assign_word_var(text_object, word_to_find, word):
    num_array = []
    i = 0
    while i < len(word_num(text_object, word_to_find)):
        num_array.append(word)
        i = i + 1 
    return(num_array)

# this function gives subject a gender
# ------------------------------------
def assign_gender(text_object, word_to_find, gender):
    num_array = []
    i = 0
    while i < len(word_num(text_object, word_to_find)):
        num_array.append(gender)
        i = i + 1 
    return(num_array)

# this function gives subject an age
# ------------------------------------
def assign_age(text_object, word_to_find, age):
    num_array = []
    i = 0
    while i < len(word_num(text_object, word_to_find)):
        num_array.append(age)
        i = i + 1 
    return(num_array)

# this function gives subject a country of origin
# -----------------------------------------------
def assign_country(text_object, word_to_find, country):
    num_array = []
    i = 0
    while i < len(word_num(text_object, word_to_find)):
        num_array.append(country)
        i = i + 1 
    return(num_array)

# this function runs all the previous functions, 
# writes the result to a csv, and returns a 
# Pandas dataframe of the information we want.
# # -------------------------------------------- 
def main_part(file_name, current_path, dataframe_path, word_to_find, subject_ID_num, gender, age, country):
    text_object = text_grabber(current_path, file_name)

    num = np.array(word_num(text_object, word_to_find))

    phrase_before = np.array(get_before(text_object, word_to_find))

    phrase_after = np.array(get_after(text_object, word_to_find))

    word_before = np.array(get_word_before(text_object, word_to_find))

    word_after = np.array(get_word_after(text_object, word_to_find))

    id_num = np.array(assign_id_number(text_object, word_to_find, subject_ID_num))

    word_var = np.array(assign_word_var(text_object, word_to_find, word_to_find))

    gender = np.array(assign_gender(text_object, word_to_find, gender))

    age = np.array(assign_age(text_object, word_to_find, age))

    country = np.array(assign_country(text_object, word_to_find, country))

    df = pd.DataFrame({"Word_Examined":word_var, "Subject_ID_Num":id_num, "Gender":gender, "Age":age, "Country":country, "Word_Appearance_Num_This_Interview":num, 
    "Phrase_Before":phrase_before, "Phrase_After":phrase_after, "Word_Before":word_before, "Word_After":word_after})

    print("The word", "\"", word_to_find, "\"", "appeared", word_counter(text_object, word_to_find), "total times in this subject's interview.")
    print(len(word_num(text_object, word_to_find)), "uses of the word were selected (based on whether they were its first appearance on a line).")
    print("A DataFrame containing this information was written to \n\"", current_path, "\" with the name", word_to_find + "_" + str(subject_ID_num) + ".csv", ".\n")
    
    df.to_csv(os.path.expanduser(dataframe_path + word_to_find + "_" + str(subject_ID_num) + ".csv"), index = False)

    return(df)

df = main_part("1718.txt", current_path, dataframe_path, "uhm", "1718", "F", 18, "United States")