Guarantee evaluation order for izip(). Document its creative uses and its limitations.

This commit is contained in:
Raymond Hettinger 2006-03-26 01:41:25 +00:00
parent 1818ed705b
commit a531e5b84c
1 changed files with 23 additions and 1 deletions

View File

@ -276,13 +276,30 @@ by functions or loops that truncate the stream.
def izip(*iterables):
iterables = map(iter, iterables)
while iterables:
result = [i.next() for i in iterables]
result = [it.next() for it in iterables]
yield tuple(result)
\end{verbatim}
\versionchanged[When no iterables are specified, returns a zero length
iterator instead of raising a \exception{TypeError}
exception]{2.4}
Note, the left-to-right evaluation order of the iterables is guaranteed.
This makes possible an idiom for clustering a data series into n-length
groups using \samp{izip(*[iter(s)]*n)}. For data that doesn't fit
n-length groups exactly, the last tuple can be pre-padded with fill
values using \samp{izip(*[chain(s, [None]*(n-1))]*n)}.
Note, when \function{izip()} is used with unequal length inputs, subsequent
iteration over the longer iterables cannot reliably be continued after
\function{izip()} terminates. Potentially, up to one entry will be missing
from each of the left-over iterables. This occurs because a value is fetched
from each iterator in-turn, but the process ends when one of the iterators
terminates. This leaves the last fetched values in limbo (they cannot be
returned in a final, incomplete tuple and they are cannot be pushed back
into the iterator for retrieval with \code{it.next()}). In general,
\function{izip()} should only be used with unequal length inputs when you
don't care about trailing, unmatched values from the longer iterables.
\end{funcdesc}
\begin{funcdesc}{repeat}{object\optional{, times}}
@ -518,4 +535,9 @@ def pairwise(iterable):
pass
return izip(a, b)
def grouper(n, iterable, padvalue=None):
"grouper(3, 'abcdefg', 'x') --> ('a','b','c'), ('d','e','f'), ('g','x','x')"
return izip(*[chain(iterable, repeat(padvalue, n-1))]*n)
\end{verbatim}