Extracting data from MS Word - python

I am looking for a way to extract / scrape data from Word files into a database. Our corporate procedures have Minutes of Meetings with clients documented in MS Word files, mostly due to history and inertia.
I want to be able to pull the action items from these meeting minutes into a database so that we can access them from a web-interface, turn them into tasks and update them as they are completed.
Which is the best way to do this:
VBA macro from inside Word to create CSV and then upload to the DB?
VBA macro in Word with connection to DB (how does one connect to MySQL from VBA?)
Python script via win32com then upload to DB?
The last one is attractive to me as the web-interface is being built with Django, but I've never used win32com or tried scripting Word from python.
EDIT: I've started extracting the text with VBA because it makes it a little easier to deal with the Word Object Model. I am having a problem though - all the text is in Tables, and when I pull the strings out of the CELLS I want, I get a strange little box character at the end of each string. My code looks like:
sFile = "D:\temp\output.txt"
fnum = FreeFile
Open sFile For Output As #fnum
num_rows = Application.ActiveDocument.Tables(2).Rows.Count
For n = 1 To num_rows
Descr = Application.ActiveDocument.Tables(2).Cell(n, 2).Range.Text
Assign = Application.ActiveDocument.Tables(2).Cell(n, 3).Range.Text
Target = Application.ActiveDocument.Tables(2).Cell(n, 4).Range.Text
If Target = "" Then
ExportText = ""
Else
ExportText = Descr & Chr(44) & Assign & Chr(44) & _
Target & Chr(13) & Chr(10)
Print #fnum, ExportText
End If
Next n
Close #fnum
What's up with the little control character box? Is some kind of character code coming across from Word?

Word has a little marker thingy that it puts at the end of every cell of text in a table.
It is used just like an end-of-paragraph marker in paragraphs: to store the formatting for the entire paragraph.
Just use the Left() function to strip it out, i.e.
Left(Target, Len(Target)-1))
By the way, instead of
num_rows = Application.ActiveDocument.Tables(2).Rows.Count
For n = 1 To num_rows
Descr = Application.ActiveDocument.Tables(2).Cell(n, 2).Range.Text
Try this:
For Each row in Application.ActiveDocument.Tables(2).Rows
Descr = row.Cells(2).Range.Text

Well, I've never scripted Word, but it's pretty easy to do simple stuff with win32com. Something like:
from win32com.client import Dispatch
word = Dispatch('Word.Application')
doc = word.Open('d:\\stuff\\myfile.doc')
doc.SaveAs(FileName='d:\\stuff\\text\\myfile.txt', FileFormat=?) # not sure what to use for ?
This is untested, but I think something like that will just open the file and save it as plain text (provided you can find the right fileformat) – you could then read the text into python and manipulate it from there. There is probably a way to grab the contents of the file directly, too, but I don't know it off hand; documentation can be hard to find, but if you've got VBA docs or experience, you should be able to carry them across.
Have a look at this post from a while ago: http://mail.python.org/pipermail/python-list/2002-October/168785.html Scroll down to COMTools.py; there's some good examples there.
You can also run makepy.py (part of the pythonwin distribution) to generate python "signatures" for the COM functions available, and then look through it as a kind of documentation.

You could use OpenOffice. It can open word files, and also can run python macros.

I'd say look at the related questions on the right -->
The top one seems to have some good ideas for going the python route.

how about saving the file as xml. then using python or something else and pull the data out of word and into the database.

It is possible to programmatically save a Word document as HTML and to import the table(s) contained into Access. This requires very little effort.

Related

Extract text from a config file [duplicate]

This question already has answers here:
Parse key value pairs in a text file
(7 answers)
Closed 1 year ago.
I'm using a config file to inform my Python script of a few key-values, for use in authenticating the user against a website.
I have three variables: the URL, the user name, and the API token.
I've created a config file with each key on a different line, so:
url:<url string>
auth_user:<user name>
auth_token:<API token>
I want to be able to extract the text after the key words into variables, also stripping any "\n" that exist at the end of the line. Currently I'm doing this, and it works but seems clumsy:
with open(argv[1], mode='r') as config_file:
lines = config_file.readlines()
for line in lines:
url_match = match('jira_url:', line)
if url_match:
jira_url = line[9:].split("\n")[0]
user_match = match('auth_user:', line)
if user_match:
auth_user = line[10:].split("\n")[0]
token_match = match('auth_token', line)
if token_match:
auth_token = line[11:].split("\n")[0]
Can anybody suggest a more elegant solution? Specifically it's the ... = line[10:].split("\n")[0] lines that seem clunky to me.
I'm also slightly confused why I can't reuse my match object within the for loop, and have to create new match objects for each config item.
you could use a .yml file and read values with yaml.load() function:
import yaml
with open('settings.yml') as file:
settings = yaml.load(file, Loader=yaml.FullLoader)
now you can access elements like settings["url"] and so on
If the format is always <tag>:<value> you can easily parse it by splitting the line at the colon and filling up a custom dictionary:
config_file = open(filename,"r")
lines = config_file.readlines()
config_file.close()
settings = dict()
for l in lines:
elements = l[:-1].split(':')
settings[elements[0]] = ':'.join(elements[1:])
So, you get a dictionary that has the tags as keys and the values as values. You can then just refer to these dictionary entries in your pogram.
(e.g.: if you need the auth_token, just call settings["auth_token"]
if you can add 1 line for config file, configparser is good choice
https://docs.python.org/3/library/configparser.html
[1] config file : 1.cfg
[DEFAULT] # configparser's config file need section name
url:<url string>
auth_user:<user name>
auth_token:<API token>
[2] python scripts
import configparser
config = configparser.ConfigParser()
config.read('1.cfg')
print(config.get('DEFAULT','url'))
print(config.get('DEFAULT','auth_user'))
print(config.get('DEFAULT','auth_token'))
[3] output
<url string>
<user name>
<API token>
also configparser's methods is useful
whey you can't guarantee config file is always complete
You have a couple of great answers already, but I wanted to step back and provide some guidance on how you might approach these problems in the future. Getting quick answers sometimes prevents you from understanding how those people knew about the answers in the first place.
When you zoom out, the first thing that strikes me is that your task is to provide config, using a file, to your program. Software has the remarkable property of solve-once, use-anywhere. Config files have been a problem worth solving for at least 40 years, so you can bet your bottom dollar you don't need to solve this yourself. And already-solved means someone has already figured out all the little off-by-one and edge-case dramas like stripping line endings and dealing with expected input. The challenge of course, is knowing what solution already exists. If you haven't spent 40 years peeling back the covers of computers to see how they tick, it's difficult to "just know". So you might have a poke around on Google for "config file format" or something.
That would lead you to one of the most prevalent config file systems on the planet - the INI file. Just as useful now as it was 30 years ago, and as a bonus, looks not too dissimilar to your example config file. Then you might search for "read INI file in Python" or something, and come across configparser and you're basically done.
Or you might see that sometime in the last 30 years, YAML became the more trendy option, and wouldn't you know it, PyYAML will do most of the work for you.
But none of this gets you any better at using Python to extract from text files in general. So zooming in a bit, you want to know how to extract parts of lines in a text file. Again, this problem is an age-old problem, and if you were to learn about this problem (rather than just be handed the solution), you would learn that this is called parsing and often involves tokenisation. If you do some research on, say "parsing a text file in python" for example, you would learn about the general techniques that work regardless of the language, such as looping over lines and splitting each one in turn.
Zooming in one more step closer, you're looking to strip the new line off the end of the string so it doesn't get included in your value. Once again, this ain't a new problem, and with the right keywords you could dig up the well-trodden solutions. This is often called "chomping" or "stripping", and with some careful search terms, you'd find rstrip() and friends, and not have to do awkward things like splitting on the '\n' character.
Your final question is about re-using the match object. This is much harder to research. But again, the "solution" wont necessarily show you where you went wrong. What you need to keep in mind is that the statements in the for loop are sequential. To think them through you should literally execute them in your mind, one after one, and imagine what's happening. Each time you call match, it either returns None or a Match object. You never use the object, except to check for truthiness in the if statement. And next time you call match, you do so with different arguments so you get a new Match object (or None). Therefore, you don't need to keep the object around at all. You can simply do:
if match('jira_url:', line):
jira_url = line[9:].split("\n")[0]
if match('auth_user:', line):
auth_user = line[10:].split("\n")[0]
and so on. Not only that, if the first if triggered then you don't need to bother calling match again - it will certainly not trigger any of other matches for the same line. So you could do:
if match('jira_url:', line):
jira_url = line[9:].rstrip()
elif match('auth_user:', line):
auth_user = line[10:].rstrip()
and so on.
But then you can start to think - why bother doing all these matches on the colon, only to then manually split the string at the colon afterwards? You could just do:
tokens = line.rstrip().split(':')
if token[0] == 'jira_url':
jira_url = token[1]
elif token[0] == 'auth_user':
auth_user = token[1]
If you keep making these improvements (and there's lots more to make!), eventually you'll end up re-writing configparse, but at least you'll have learned why it's often a good idea to use an existing library where practical!

Extracting Fasta Moonlight Protein Sequences with Python

I want to extract the FASTA files that have the aminoacid sequence from the Moonlighting Protein Database ( www.moonlightingproteins.org/results.php?search_text= ) via Python, since it's an iterative process, which I'd rather learn how to program than manually do it, b/c come on, we're in 2016. The problem is I don´t know how to write the code, because I'm a rookie programmer :( . The basic pseudocode would be:
for protein_name in site: www.moonlightingproteins.org/results.php?search_text=:
go to the uniprot option
download the fasta file
store it in a .txt file inside a given folder
Thanks in advance!
I would strongly suggest to ask the authors for the database. From the FAQ:
I would like to use the MoonProt database in a project to analyze the
amino acid sequences or structures using bioinformatics.
Please contact us at bioinformatics#moonlightingproteins.org if you are
interested in using MoonProt database for analysis of sequences and/or
structures of moonlighting proteins.
Assuming you find something interesting, how are you going to cite it in your paper or your thesis?
"The sequences were scraped from a public webpage without the consent of the authors". Much better to give credit to the original researchers.
That's a good introduction to scraping
But back to your your original question.
import requests
from lxml import html
#let's download one protein at a time, change 3 to any other number
page = requests.get('http://www.moonlightingproteins.org/detail.php?id=3')
#convert the html document to something we can parse in Python
tree = html.fromstring(page.content)
#get all table cells
cells = tree.xpath('//td')
for i, cell in enumerate(cells):
if cell.text:
#if we get something which looks like a FASTA sequence, print it
if cell.text.startswith('>'):
print(cell.text)
#if we find a table cell which has UniProt in it
#let's print the link from the next cell
if 'UniProt' in cell.text_content():
if cells[i + 1].find('a') is not None and 'href' in cells[i + 1].find('a').attrib:
print(cells[i + 1].find('a').attrib['href'])

Python: Writing text to a 2003 Word Doc in a specific place on the page

I'm using Python 2.7, Windows 7, and Word 2003. Those three cannot change (well except for maybe the python version). I work in Law and the attorneys have roughly 3 boiler plate objections (just a large piece of text, maybe 5 paragraphs) that need to be inserted into a word document at a specific spot. Now instead of going through and copying and pasting the objection where its needed, my idea is for the user to go through the document adding a special word/phrase (place holder if you will) that wont be found anywhere in the document. Then run some code and have python fill in the rest. Maybe not the cleverest way to go about it, but I'm a noob. I've been practicing with a test page and inserted the below text as place holders (the extra "o" stands for objection)
oone
otwo
othree
Below is what I have so far. I have two questions
Do you have any other suggestions to go about this?
My code does insert the string in the correct order, but the formatting goes out the window and it writes in my string 6 times instead of 1. How can I resolve the formatting issue so it simply writes the text into the spot the place holder is at?
import sys
import fileinput
f = open('work.doc', 'r+')
obj1 = "oone"
obj2 = "otwo"
obj3 = "othree"
for line in fileinput.input('work.doc'):
if obj1 in line:
f.write("Objection 1")
elif obj2 in line:
f.write("Objection 2")
elif obj3 in line:
f.write("Objection 3")
else:
f.write("No Objection")
f.close
You could use python-uno to load the document into OpenOffice and manipulate it using the UNO interface. There is some example code on the site I just linked to which can get you started.

Printing in Duplexpage a word document

I am trying to automate the task of printing two copies at double page of ~30 Word document (*.doc). I want to send the program converted to .exe (I plan it just for Windows computers) using py2exe. I know that I can manually check the options but I will not be able to do so on the 20 or so computer where it will be used, as well as I cannot install in this computers new software (That's why I want to convert it into .exe).
I copied this solution to print, but I can't adapt it to do what I want:
from win32com import client
import time
word = client.Dispatch("Word.Application")
filename=input("What files do you want to print?")
def printWordDocument(filename):
"""Given a name of a file prints it. TODO: Add double page."""
word.Documents.Open(filename)
word.ActiveDocument.PrintOut()
time.sleep(2)
word.ActiveDocument.Close()
word.Quit()
I couldn't find any option to print in double pages, or at least automatically, the only option of double page of PrintOut method is ManualDuplexPrint which in the documentation says: "True to print a two-sided document on a printer without a duplex printing kit.", but I don't want to make it even easier to print all the set of documents. And make a program portable to other computers, without modifying the Word documents (I don't create them).
Any other way to do it? Or any other option to do it?
UPDATE
I am not able to code in visual-basic (yet), but if I get a template or some hints I think I will manage to make something adapted to my conditions.
I have ended doing a macro, but this just works for my own computer and not for all the computers where should work.
Sub Test()
'
' Test Macro
' Print in double page and 2 copies
'
ActivePrinter = "Xerox WC 24 PCL"
Application.PrintOut FileName:="", Range:=wdPrintAllDocument, Item:= _
wdPrintDocumentWithMarkup, Copies:=2, Pages:="", PageType:= _
wdPrintAllPages, Collate:=True, Background:=True, PrintToFile:=False, _
PrintZoomColumn:=0, PrintZoomRow:=0, PrintZoomPaperWidth:=0, _
PrintZoomPaperHeight:=0
End Sub

Python: Matching & Stripping port number from socket data

I have data coming in to a python server via a socket. Within this data is the string '<port>80</port>' or which ever port is being used.
I wish to extract the port number into a variable. The data coming in is not XML, I just used the tag approach to identifying data for future XML use if needed. I do not wish to use an XML python library, but simply use something like regexp and strings.
What would you recommend is the best way to match and strip this data?
I am currently using this code with no luck:
p = re.compile('<port>\w</port>')
m = p.search(data)
print m
Thank you :)
Regex can't parse XML and shouldn't be used to parse fake XML. You should do one of
Use a serialization method that is nicer to work with to start with, such as JSON or an ini file with the ConfigParser module.
Really use XML and not something that just sort of looks like XML and really parse it with something like lxml.etree.
Just store the number in a file if this is the entirety of your configuration. This solution isn't really easier than just using JSON or something, but it's better than the current one.
Implementing a bad solution now for future needs that you have no way of defining or accurately predicting is always a bad approach. You will be kept busy enough trying to write and maintain software now that there is no good reason to try to satisfy unknown future needs. I have never seen a case where "I'll put this in for later" has led to less headache later on, especially when I put it in by doing something completely wrong. YAGNI!
As to what's wrong with your snippet other than using an entirely wrong approach, angled brackets have a meaning in regex.
Though Mike Graham is correct, using regex for xml is not 'recommended', the following will work:
(I have defined searchType as 'd' for numerals)
searchStr = 'port'
if searchType == 'd':
retPattern = '(<%s>)(\d+)(</%s>)'
else:
retPattern = '(<%s>)(.+?)(</%s>)'
searchPattern = re.compile(retPattern % (searchStr, searchStr))
found = searchPattern.search(searchStr)
retVal = found.group(2)
(note the complete lack of error checking, that is left as an exercise for the user)

Categories

Resources