2:14 p.m.
Hello,
Is there a way within hansl to "scan" a given folder and give back the subfolder
names, e.g. as strings? Non-recursive would be enough.
Background: I am still working on the data-input script and changed the format of my
input. If I could scan folders, I could get rid of some additional input information.
Best
Frederik
8:38 a.m.
Am 16.10.2017 um 16:14 schrieb Schaff, Frederik:
Hello,
Is there a way within hansl to “scan” a given folder and give back the
subfolder names, e.g. as strings? Non-recursive would be enough.
Spontaneously I'd say this is not possible "within" hansl. However, with
embedded shell commands (look for '!' and 'shell') it would "of
course"
be feasible. You can also use python or similar programs instead of a
shell in a narrow sense.
hth,
sven
9:13 a.m.
On Tue, 17 Oct 2017, Sven Schreiber wrote:
Am 16.10.2017 um 16:14 schrieb Schaff, Frederik:
> Hello,
>
> Is there a way within hansl to “scan” a given folder and give back the
> subfolder names, e.g. as strings? Non-recursive would be enough.
Spontaneously I'd say this is not possible "within" hansl. However, with
embedded shell commands (look for '!' and 'shell') it would "of
course" be
feasible. You can also use python or similar programs instead of a shell in a
narrow sense.
Very nice idea. If you have Python installed, you may find this useful:
<hansl>
function strings subdirs(string path)
outfile s --buffer --quiet
foreign language=python
from os import walk
a = next(walk("@path"))[1]
for d in a:
print(d)
end foreign
outfile --close
l = ""
strings ret = null
loop while getline(s, l) --quiet
ret += l
endloop
return ret
end function
path = "@dotdir"
a = subdirs(path)
print a
</hansl>
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
10:44 a.m.
Am 17.10.2017 um 11:13 schrieb Riccardo (Jack) Lucchetti:
Very nice idea. If you have Python installed, you may find this
useful:
<hansl>
function strings subdirs(string path)
outfile s --buffer --quiet
foreign language=python
Well, for once this is a justification that it was good to use
language=python instead of language=numpy.
thanks,
sven
6:11 p.m.
Many thanks and sorry for the delay, so I will make use of python. I just thought there
might be a gretl native option ;). For the record: The option by Riccardo does not work
for me, but I am working on it...
Best
Frederik
-----Ursprüngliche Nachricht-----
Von: gretl-users-bounces(a)lists.wfu.edu [mailto:gretl-users-
bounces(a)lists.wfu.edu] Im Auftrag von Sven Schreiber
Gesendet: Dienstag, 17. Oktober 2017 12:45
An: gretl-users(a)lists.wfu.edu
Betreff: Re: [Gretl-users] Scan folder
Am 17.10.2017 um 11:13 schrieb Riccardo (Jack) Lucchetti:
> Very nice idea. If you have Python installed, you may find this useful:
>
> <hansl>
> function strings subdirs(string path)
> outfile s --buffer --quiet
> foreign language=python
Well, for once this is a justification that it was good to use language=python
instead of language=numpy.
thanks,
sven
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
6:25 p.m.
Am 20.10.2017 um 20:11 schrieb Schaff, Frederik:
Many thanks and sorry for the delay, so I will make use of python. I
just thought there might be a gretl native option ;). For the record:
The option by Riccardo does not work for me, but I am working on
it...
What's your python version? For example, this part of Jack's embedded
python code: "print(d)" looks like Python3, so I think it wouldn't work
in Python 2.x.
If that's not it, you might post some details about the problem anyway.
thanks,
sven
6:14 a.m.
What's your python version? For example, this part of Jack's
embedded
python code: "print(d)" looks like Python3, so I think it wouldn't work in
Python 2.x.
If that's not it, you might post some details about the problem anyway.
thanks,
sven
I am using python 3. Here is the gretl output, using the script of Jack (i.e. with dotdir
as path - the problem is general, however).
----
Array of strings, length 4
[1] " File "C:\Users\Frederik\AppData\Roaming\gretl\gretltmp.py", line
3"
[2] " a =
next(walk("C:\Users\Frederik\AppData\Roaming\gretl"))[1]"
[3] " ^"
[4] "SyntaxError: (unicode error) 'unicodeescape' codec can't decode
bytes..."
---
Best
Frederik
9:42 a.m.
On Wed, 25 Oct 2017, Schaff, Frederik wrote:
> What's your python version? For example, this part of
Jack's embedded
> python code: "print(d)" looks like Python3, so I think it wouldn't work
in
> Python 2.x.
> If that's not it, you might post some details about the problem anyway.
>
> thanks,
> sven
I am using python 3. Here is the gretl output, using the script of Jack (i.e. with dotdir
as path - the problem is general, however).
----
Array of strings, length 4
[1] " File "C:\Users\Frederik\AppData\Roaming\gretl\gretltmp.py", line
3"
[2] " a =
next(walk("C:\Users\Frederik\AppData\Roaming\gretl"))[1]"
[3] " ^"
[4] "SyntaxError: (unicode error) 'unicodeescape' codec can't decode
bytes..."
---
Gaaaaaaaaah, that's because "\U" means "Unicode" in Python. If the
string
containing tha path is "raw-ified", the problem should go. Could you try
substituting
<python>
a = next(walk("@path"))[1]
</python>
with
<python>
a = next(walk(r"@path"))[1]
</python>
and see what happens? (The prepended "r" means "raw")
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
10:53 a.m.
Gaaaaaaaaah, that's because "\U" means "Unicode" in Python. If
the string
containing tha path is "raw-ified", the problem should go. Could you try
substituting
<python>
a = next(walk("@path"))[1] </python>
with
<python>
a = next(walk(r"@path"))[1] </python>
and see what happens? (The prepended "r" means "raw")
That does the trick, thanks! Although again I encountered one of my "print"
problems. 'print a' produces " Array of strings, length 11"; I tried a
simple loop construct to print the content of a:
<hansl>
loop for i=1..nelem(a)
print a[i]
endloop
<\hansl>
but that does not work. Instead I have to assign 'z=a[i]' first and then
'print z'.
Best
Frederik
12:04 p.m.
Am 25.10.2017 um 12:53 schrieb Schaff, Frederik:
> and see what happens? (The prepended "r" means
"raw")
(I'm not sure this would work on Python-2 with non-Ascii filenames,
perhaps ur"@path" might be needed there, but cannot test right now, and
Frederik had Python-3 anyway.)
loop for i=1..nelem(a)
Actually, you should omit the 'for' in this simple variant, just
'loop i=1..nelem(a)'
But it's tolerated by gretl.
print a[i]
Yes, known limitation of print, by design. Use eval, or:
printf "%s\n", a[i]
cheers,
sven
2:27 p.m.
On Wed, 25 Oct 2017, Schaff, Frederik wrote:
I am using python 3. Here is the gretl output, using the script of
Jack (i.e. with dotdir as path - the problem is general, however).
----
Array of strings, length 4
[1] " File "C:\Users\Frederik\AppData\Roaming\gretl\gretltmp.py", line
3"
[2] " a =
next(walk("C:\Users\Frederik\AppData\Roaming\gretl"))[1]"
[3] " ^"
[4] "SyntaxError: (unicode error) 'unicodeescape' codec can't decode
bytes..."
---
Microsoft's backslash path-separation is a minefield for text
processing. You're much better off using forward slashes (which work
fine). If you get a backslash-separated path from gretl, or elsewhere,
you can fix it using strsub:
string path = $dotdir
string bslash = "\"
path = strsub(path, bslash, "/")
Allin
2369
days inactive
2378
days old
10 comments
4 participants
participants (4)
-
Allin Cottrell -
Riccardo (Jack) Lucchetti -
Schaff, Frederik -
Sven Schreiber