I'm trying to write a vim plugin that uses a Python code block inside of it. I would like to get the full path of myvim.vim (/home/myusername/.vim/bundle/myvim/plugin/myvim.vim) inside of my python code block. Unfortunately you can't get the path by using __file__ as in a .py file. I can't use vim.command(':pwd') either because that just prints the path of the location where the plugin function is called from.
myvim.vim
function! Myvim()
python << EOF
import vim
vim_path = "full myvim.vim path here"
print vim_path
EOF
endfunction
EDIT
#actionshrimp, I'm trying this:
myvim.vim
function! Myvim()
let s:curfile = expand("<sfile>")
let s:curfiledir = fnamemodify(s:curfile, ":h")
python << EOF
import vim
py vim_path = vim.eval('expand("<sfile>")')
print vim_path
EOF
endfunction
You can use <sfile> to get the path of the currently executing vimscript, like so:
let s:curfile = expand("<sfile>")
let s:curfiledir = fnamemodify(s:curfile, ":h")
To pass that to python you should be able to use:
py vim_path = vim.eval('expand("<sfile>")')
or if you've set the variable:
py vim_path = vim.eval('s:curfile')
For clarity here's a full example (saved as 'D:\tmp\test.vim'):
python << EOF
import vim
vim_path = vim.eval('expand("<sfile>")')
print vim_path
EOF
When I have it open and type :so % it shows 'D:\tmp\test.vim' at the bottom.
Related
I'm facing quite simple problem, yet still not able to figure out what in particular causes that, and - more importantly - how to solve it.
I'm currently on Linux, and I would like to run an python app (script) from a Java application. The script is on PATH, thus I really would like to utilise that (avoid using absolute path to the script), if possible.
However, all I tried resulted in various forms of "File does not exist".
Sample
As a demonstration, I've tried to run one python3-based app (meld)
and one binary-built app (ls) for comparison:
$ which ls
/usr/bin/ls
$ file /usr/bin/ls
/usr/bin/ls: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, BuildID[sha1]=[...], for GNU/Linux 3.2.0, stripped
$ which meld
/usr/bin/meld
$ file /usr/bin/meld
/usr/bin/meld: Python script, UTF-8 Unicode text executable
$ head -n1 /usr/bin/meld
#!/usr/bin/python3
Nextly, I've created simple Java main, which tries several ways how to start theese:
package cz.martlin.processes;
import java.io.File;
import java.io.IOException;
import java.lang.ProcessBuilder.Redirect;
import java.util.List;
import java.util.stream.Collectors;
public class ProcessPlaying {
public static void main(String[] args) {
List<List<String>> commands = List.of(
List.of("ls"),
List.of("/usr/bin/ls"),
List.of("meld"),
List.of("/usr/bin/meld"),
List.of("python3", "/usr/bin/meld"),
List.of("/usr/bin/python3", "/usr/bin/meld"),
List.of("sh", "-c", "meld"),
List.of("sh", "-c", "/usr/bin/meld"),
List.of("sh", "-c", "python3 /usr/bin/meld")
);
for (List<String> command : commands) {
run(command);
}
}
private static void run(List<String> command) {
System.out.println("Running: " + command);
//String executable = command.get(0);
//boolean exists = new File(executable).exists();
//System.out.println("Exists the " + executable + " ? " + exists);
try {
ProcessBuilder pb = new ProcessBuilder(command);
pb.redirectError(Redirect.INHERIT);
Process proc = pb.start();
// Process proc = Runtime.getRuntime().exec(command.stream().collect(Collectors.joining(" ")));
int code = proc.waitFor();
System.out.println("OK, return code: " + code);
} catch (IOException e) {
System.out.println("Failed to start: " + e.toString());
} catch (InterruptedException e) {
System.out.println("Failed to await: " + e.toString());
}
System.out.println();
}
}
Here are the results:
Running: [ls]
OK, return code: 0
Running: [/usr/bin/ls]
OK, return code: 0
Running: [meld]
Failed to start: java.io.IOException: Cannot run program "meld": error=2, Directory or file doesn't exist
Running: [/usr/bin/meld]
Failed to start: java.io.IOException: Cannot run program "/usr/bin/meld": error=2, Directory or file doesn't exist
Running: [python3, /usr/bin/meld]
python3: can't open file '/usr/bin/meld': [Errno 2] No such file or directory
OK, return code: 2
Running: [/usr/bin/python3, /usr/bin/meld]
/usr/bin/python3: can't open file '/usr/bin/meld': [Errno 2] No such file or directory
OK, return code: 2
Running: [sh, -c, meld]
sh: line 1: meld: command not found
OK, return code: 127
Running: [sh, -c, /usr/bin/meld]
sh: line 1: /usr/bin/meld: Directory or file doesn't exist
OK, return code: 127
Running: [sh, -c, python3 /usr/bin/meld]
python3: can't open file '/usr/bin/meld': [Errno 2] No such file or directory
OK, return code: 2
To sum it up:
all of the commands tried works when executed directly from the shell (either sh or bash)
executing the binary works either by the full path (/usr/bin/ls) or by just the name (ls)
executing the python script the same way doesn't work neither way
when trying to run the python3 interpreter and populate the script path as an argument to the python, now the python yields the script file doesn't exist
trying to populate it as a command to the brand new shell didn't help either
I've tried to use the Runtime#exec (based on this comment: https://stackoverflow.com/a/36783743/3797793) to start the process (both the exec(String) and exec(String[] forms), but no sucess (none of the listed commands did actually execute).
Thus, my question is/are:
What do I understand wrong?
How to start the Python script from Java?
Would some small (ba)sh script wrapper do the job?
Since it's python3-based, Jython wouldn't help here (because the latest one is 2.7.*), would it?
Further requirements:
As mentioned, I would like to be able to avoid using full path to the Python script
Also, I would like to have platform independant solution (at least Linux and Windows compatible)
Thanks in advance
I want to run another file from my main.py and also set the value of a variable in the main.py and "export it", so I can use it in the other file.
But I absolutely don´t know how to do that.
So that I can set a 'var = x[3]' in the main file and kinda "export" the variable into the file that I am opening like 'import otherfile.py(var = x[3])'
Thanks in advance
To import any variable from main file to otherfile you can use,
from main import*
Y = var
That's how you can set var in main.py file and import or use it in otherfile.py.
You can use sys.argv to get arguments if you run the file from the command line
so if you have the file:
import sys
print(sys.argv[0])
then open a command line and type python file.py hello it will print 'hello'.
As well, you can use os.system() to run command line functions from python
so your first file could be:
import os
var = 'test'
os.system(f"python otherfile.py {var}")
and otherfile.py could be:
import sys
print(sys.argv[0])
then if you run the first file, a python window will open and print 'test'.
Hope this was what you were looking for
I have a problem with converting docx to pdf files in my script.
At first I tried to use a pure php-based solution, described here:
https://stackoverflow.com/a/20035739/12812601
Unfortunatelly this does not work (it creates an empty com object, then throws a fatal error).
So I've tried to use a python script to do this.
I use a great script from here:
https://stackoverflow.com/a/20035739/12812601
So here is a problem.
The Python script standalone (run via a command line) works just fine, and saves the converted pdf. Unfortunatelly when I try to call it via PHP it can't save a converted file.
PHP scripts can create and write files oin the same directory without any problem
This supposed to be a local configuration, so I do not care about any portability
Scripts:
*******PHP*******
<?php
//Script only for testing Python calls, tried different methods
error_reporting(E_ALL);
echo '<h1>Begin</h1>';
echo '<h2>Before call</h2>';
exec ('python dp.py');
echo '<h2>After exec call</h2>';
system('python dp.py');
echo '<h2>After Sys Call</h2>';
passthru('python dp.py');
echo '<h2>After Pass Call</h2>';
$w = get_current_user();
var_dump($w);
?>
*****Python*****
import sys
import os
import comtypes.client
import win32com.client
wdFormatPDF = 17
#static file names for testing
in_file = 'C:\\Users\\fake_user\\OneDrive\\Stuff\\f1.docx'
out_file = 'C:\\Users\\fake_user\\OneDrive\\Stuff\\f3.pdf'
print('BEGIN<br>\n')
word = win32com.client.Dispatch('Word.Application')
word.Visible = False
doc = word.Documents.Open(in_file)
print('\nOpened Docx\n<br>')
print(in_file);
doc.SaveAs(out_file, FileFormat=wdFormatPDF)
print('\nSaved\n<br>')
doc.Close()
word.Quit()
print('DONE\n')
*****Output from the browser*****
Begin
Before call
After exec call
BEGIN
Opened Docx
C:\Users\fake_user\OneDrive\Stuff\f1.docx
After Sys Call
BEGIN
Opened Docx
C:\Users\fake_user\OneDrive\Stuff\f1.docx
After Pass Call
string(5) "fake_user"
System configuration
Windows 7 Professional Edition Service Pack 1
Apache/2.4.26 (Win32)
OpenSSL/1.0.2l
PHP/7.1.7
Python 3.8.1
I tried to run Apache both as a system service and as a user who owns the OneDrive (name changed to "fake_user" here), so it shouldn't be a permissions issue (I think)
Any help appreciated
I want to use syslog-ng to receive netgear log, and use python script process.
But syslog-ng didn't run the python script.
syslog-ng.config
#version:3.2
options {
flush_lines (0);
time_reopen (10);
log_fifo_size (1000);
long_hostnames (off);
use_dns (no);
use_fqdn (no);
create_dirs (no);
keep_hostname (yes);
};
source s_sys {
udp(ip(0.0.0.0) port(514));
};
destination d_python{
program("/usr/local/bin/python /opt/syslog.py");
#program("/bin/echo 'haha' >> /tmp/test");
};
log { source(s_sys); destination(d_python);};
and python script like this
#!/usr/local/bin/python
#coding:utf8
import os
import sys
import datetime
f = open('/var/log/pnet.log', 'a')
f.write('\nstart\n')
f.write('args\n')
f.write('%s\n' % sys.argv)
if not sys.stdin.isatty():
f.write('stdin\n')
f.write('%s\n' % date.date.now().isoformat() )
tty_read = sys.stdin.read()
f.write("'''\n")
f.write(tty_read)
f.write("\n'''\n")
f.write('end\n')
f.close()
The script is already 777
Even I change my config to use 'echo' directly pipe into a file, didn't write a word too...
So...why?
silly question, but do you have incoming logs? If you use a simple file destination instead of the program, do you receive logs? If not, the problem is not in the program destination.
Also, try changing the flush_lines (0); option to 1 to see if it helps.
Regards,
Robert Fekete
I could show you my code for reference:
my syslog-ng.conf :
source s_test{
file("/home/test/in.log" follow-freq(1) flags(no-parse));
};
destination d_test{
program ("/home/test/out.py" flush_lines(1) flags(no_multi_line));
};
log {
source(s_test);
destination(d_test);
flags(flow-control);
};
my out.py:
#!/usr/bin/env python
# encoding: utf-8
import sys
while True:
line = sys.stdin.readline()
file = open("/home/test/out_from_python.log","a")
file.write(line)
file.close()
when you type echo "something" >> /home/test/in.log , there would be a new log lie in /home/test/out_from_python.log
I've looked all around Google and its archives. There are several good articles, but none seem to help me out. So I thought I'd come here for a more specific answer.
The Objective: I want to run this code on a website to get all the picture files at once. It'll save a lot of pointing and clicking.
I've got Python 2.3.5 on a Windows 7 x64 machine. It's installed in C:\Python23.
How do I get this script to "go", so to speak?
=====================================
WOW. 35k views. Seeing as how this is top result on Google, here's a useful link I found over the years:
http://learnpythonthehardway.org/book/ex1.html
For setup, see exercise 0.
=====================================
FYI: I've got zero experience with Python. Any advice would be appreciated.
As requested, here's the code I'm using:
"""
dumpimages.py
Downloads all the images on the supplied URL, and saves them to the
specified output file ("/test/" by default)
Usage:
python dumpimages.py http://example.com/ [output]
"""
from BeautifulSoup import BeautifulSoup as bs
import urlparse
from urllib2 import urlopen
from urllib import urlretrieve
import os
import sys
def main(url, out_folder="C:\asdf\"):
"""Downloads all the images at 'url' to /test/"""
soup = bs(urlopen(url))
parsed = list(urlparse.urlparse(url))
for image in soup.findAll("img"):
print "Image: %(src)s" % image
filename = image["src"].split("/")[-1]
parsed[2] = image["src"]
outpath = os.path.join(out_folder, filename)
if image["src"].lower().startswith("http"):
urlretrieve(image["src"], outpath)
else:
urlretrieve(urlparse.urlunparse(parsed), outpath)
def _usage():
print "usage: python dumpimages.py http://example.com [outpath]"
if __name__ == "__main__":
url = sys.argv[-1]
out_folder = "/test/"
if not url.lower().startswith("http"):
out_folder = sys.argv[-1]
url = sys.argv[-2]
if not url.lower().startswith("http"):
_usage()
sys.exit(-1)
main(url, out_folder)
On windows platform, you have 2 choices:
In a command line terminal, type
c:\python23\python xxxx.py
Open the python editor IDLE from the menu, and open xxxx.py, then press F5 to run it.
For your posted code, the error is at this line:
def main(url, out_folder="C:\asdf\"):
It should be:
def main(url, out_folder="C:\\asdf\\"):
Usually you can double click the .py file in Windows explorer to run it. If this doesn't work, you can create a batch file in the same directory with the following contents:
C:\python23\python YOURSCRIPTNAME.py
Then double click that batch file. Or, you can simply run that line in the command prompt while your working directory is the location of your script.
Since you seem to be on windows you can do this so python <filename.py>. Check that python's bin folder is in your PATH, or you can do c:\python23\bin\python <filename.py>. Python is an interpretive language and so you need the interpretor to run your file, much like you need java runtime to run a jar file.
use IDLE Editor {You may already have it} it has interactive shell for python and it will show you execution and result.
Your command should include the url parameter as stated in the script usage comments.
The main function has 2 parameters, url and out (which is set to a default value)
C:\python23\python "C:\PathToYourScript\SCRIPT.py" http://yoururl.com "C:\OptionalOutput\"
If you want to run .py files in Windows, Try installing Git bash
Then download python(Required Version) from python.org and install in the main c drive folder
For me, its :
"C:\Python38"
then open Git Bash and go to the respective folder where your .py file is stored :
For me, its :
File Location : "Downloads"
File Name : Train.py
So i changed my Current working Directory From "C:/User/(username)/" to "C:/User/(username)/Downloads"
then i will run the below command
" /c/Python38/python Train.py "
and it will run successfully.
But if it give the below error :
from sklearn.model_selection import train_test_split
ModuleNotFoundError: No module named 'sklearn'
Then Do not panic :
and use this command :
" /c/Python38/Scripts/pip install sklearn "
and after it has installed sklearn go back and run the previous command :
" /c/Python38/python Train.py "
and it will run successfully.
!!!!HAPPY LEARNING !!!!