如何用空白填充信息，而其余部分保持不变？

Question

I am constructing scenery for a Flight Simulator and need to figure out how to edit many lines in a text file (3,579,189 of them).

I have TextCrawler Pro, Node, Python SVN and Notepad++ as tools.

Raw, pre-edit portion:

POLYGON_POINT -79.750000000217,42.017498354525,0
POLYGON_POINT -79.750000000217,42.016478251402,0
POLYGON_POINT -79.750598748133,42.017193264943,0
POLYGON_POINT -79.750000000217,42.017498354525,0


POLYGON_POINT -79.750000000217,42.085882815878,0
POLYGON_POINT -79.750000000217,42.082008734634,0
POLYGON_POINT -79.751045507507,42.082126409633,0
POLYGON_POINT -79.750281907508,42.083166574215,0
POLYGON_POINT -79.750781149174,42.084212672130,0
POLYGON_POINT -79.750000000217,42.085882815878,0


POLYGON_POINT -79.750000000217,42.088955814831,0
POLYGON_POINT -79.750456566883,42.087544672125,0
POLYGON_POINT -79.751642899173,42.088273325249,0
POLYGON_POINT -79.751461052298,42.088916154415,0
POLYGON_POINT -79.750000000217,42.088955814831,0

With Notepad++'s replace function, it is easy enough to add the POLYGON_POINT line. Now I need some assistance in making it appear as so:

BEGIN_POLYGON
POLYGON_POINT -79.750000000217,42.017498354525,0
POLYGON_POINT -79.750000000217,42.016478251402,0
POLYGON_POINT -79.750598748133,42.017193264943,0
POLYGON_POINT -79.750000000217,42.017498354525,0
END_POLY
BEGIN_POLYGON
POLYGON_POINT -79.750000000217,42.085882815878,0
POLYGON_POINT -79.750000000217,42.082008734634,0
POLYGON_POINT -79.751045507507,42.082126409633,0
POLYGON_POINT -79.750281907508,42.083166574215,0
POLYGON_POINT -79.750781149174,42.084212672130,0
POLYGON_POINT -79.750000000217,42.085882815878,0
END_POLY
BEGIN_POLYGON
POLYGON_POINT -79.750000000217,42.088955814831,0
POLYGON_POINT -79.750456566883,42.087544672125,0
POLYGON_POINT -79.751642899173,42.088273325249,0
POLYGON_POINT -79.751461052298,42.088916154415,0
POLYGON_POINT -79.750000000217,42.088955814831,0

ie add BEGIN_POLYGON before each block and END_POLY after each.

How can I do this?

Answer 1

I would group the lines by being blank or not, using itertools.groupby (only taking the non-blank groups with the if k condition), and add the header/footer for each group. Then flatten the groups using itertools.chain

import itertools

with open("file.txt") as f, open("fileout.txt","w") as fw:
    fw.writelines(itertools.chain.from_iterable([["BEGIN_POLYGON\n"]+list(v)+["END_POLYGON\n"] for k,v in itertools.groupby(f,key = lambda l : bool(l.strip())) if k]))

key = lambda l : bool(l.strip())) is the grouping key: test for empty line but for line termination

this method doesn't need to read the file fully, so it's suited for very big files. It processes the file line by line so it doesn't hog the memory.

Answer 2

A quick solution using sed

cat -s file.txt |\
    sed -e 's/^$/END_POLY\nBEGIN_POLYGON/'\
    -e '1i BEGIN_POLYGON'\
    -e '$a END_POLY'

• cat -s squeezes all blank lines into one
• the first sed replace blank lines with END_POLY and BEGIN_POLYGON tags
• the second and last sed prepends and appends the remaining tags to the output