| Re: Is there a way to use C++ code in IDL [message #34193] |
Thu, 20 February 2003 12:07 |
Rick Towler
Messages: 821
Registered: August 1998
|
Senior Member |
|
|
Here is a pile of bad code for you :)
I wrote a file parser for some home grown tagged data files we have around
here. This is the gist of it altered for your file format. It is ugly but
should get you well on your way.
I still think you should do this as a dlm but this exercise will be a good
intro to IDL. Note that I have left a lot out and I make no claims that
this is the way anyone should do this.
Enjoy.
-Rick
datasets = ''
nsets = 0
;read in the batch file
openr, lun, filename, /get_lun
while not eof(lun) do begin
line = ''
readf, lun, line
;remove comments and whitespace
temp = strsplit(line, '//', /extract)
line = strtrim(temp[0],2)
;check if this is a datablock tag
dblock = strsplit(line, ':', /extract)
if (N_ELEMENTS(dblock eq 1) then begin
;not a datablock tag
case line of
'[System]':begin
while (line ne '[End]') do begin
line=''
readf, lun, line
;remove comments and whitespace
temp = strsplit(line, '//', /extract)
line = strtrim(temp[0],2)
;parse the line of text for the keyword/param pair
kwpair = strsplit(strcompress(line, /remove_all), $
'=', /extract)
keyword = kwpair[0]
case keyword of
'Name':name = FIX(kwpair[1])
else: ;skip everything else
endcase
endwhile
end
'[Data Sets]':begin
while (line ne '[End]') do begin
line=''
readf, lun, line
;remove comments and whitespace
line = strtrim((strsplit(line, '//', /extract))[0],2)
datasets = [datasets,line]
nsets = nsets + 1
endwhile
end
'[File Attributes]':begin
while (line ne '[End]') do begin
line=''
readf, lun, line
;remove comments and whitespace
line = strtrim((strsplit(line, '//', /extract))[0],2)
case line of
'[Conditions File]':begin
while (line ne '[End]') do begin
line=''
readf, lun, line
;remove comments and whitespace
line = strtrim((strsplit(line, '//', $
/extract))[0],2)
case line of
'[Properties]':begin
while (line ne '[End]') do begin
line=''
readf, lun, line
;remove comments and whitespace
line = strtrim((strsplit(line,
'//', $
/extract))[0],2)
;process keyword/data pairs
endwhile
end
'[Frame Information]':begin
end
endcase
endwhile
end
'[Saved File]':begin
end
else:;skip everything else
endcase
endwhile
end
else:;ignore everything else
endcase
endif else begin
;This is a datablock tag
;dblock[1] will be the dataset label
endelse
endwhile
free_lun, lun
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34194 is a reply to message #34193] |
Thu, 20 February 2003 11:55  |
CelticBlues
Messages: 10
Registered: February 2003
|
Junior Member |
|
|
Good stuff... thanks for the info on structs...
Ed
"Paul van Delst" <paul.vandelst@noaa.gov> wrote in message
news:3E552C46.ECC98EAB@noaa.gov...
> CelticBlues wrote:
>>
>> The file is a home grown tagged file format containing a type of image
data
>> and additional information about the conditions at the time the image
was
>> taken etc. Here is an example of what a simple file may look like:
>> ---------------------------------------------------------
>> [System]
>> Name = 1
>> [End]
>> [Data Sets] // will be at least one data set, could be as many as 3
>> DataSet1
>> [End]
>> [Files Attributes]
>> [Conditions File]
>> [Properties]
>> Name = 97918.10
>> CreateTime = Thu Sept 18 1997 at 11:21:36
>> LastAccessTime = Mon Jan 01 1601
>> LastModifiedTime = Thu Sept 18 1997 at 11:21:36
>> Size = 16882000 // bytes
>> [End]
>> [Frame information]
>> Width = 133
>> Height = 148
>> [End]
>> [End] // end [Telemetry File]
>>
>> [Saved File]
>> CreateDate = Tue Feb 11 2003 at 16:08:54
>> Width = 133
>> Height = 148
>> FrameNumber = 1
>> DataUnits = meters
>> Comment =
>> CoordinateSystem = UTM // Latlong or UTM
>> [End]
>>
>> [End] // end [Files Attributes]
>>
>> [DataBlock:DataSet1]
>> R1P1 R1P2...R1P148
>> R2P1 R2P2...R2P148
>> .
>> .
>> .
>> R133P1 R133P2...R133P148
>> [End]
>>
>> ---------------------------------------------------------
>>
>> R1P1 = Row 1 Pixel 1, R2P1 = Row 2, Pixel 2 etc.
>>
>> One additional bit of info... the datablock may not actually have CR/NL
>> after each row, i.e. it may be
>>
>> [DataBlock:DataSet1]
>> R1P1 R1P2...R1P148 R2P1 R2P2...R2P148 ... R133P1 R133P2...R133P148
>> [End]
>>
>> Ok, assume that I want to rewrite the parser in IDL... From what I
>> understand it is simple enough to read an ascii table from a file, but
what
>> about reading this type of format?
>
> The reading isn't the problem - it's the sorting of the read-in bits. :o)
>
> But, it doesn't look too difficult. Given what you know about the
allowable syntax in your
> ASCII file, you can create the structures up front, e.g.
>
> Properties = { Name: 0.0, $
> CreateTime: ' ', $
> LastAccessTime: ' ', $
> LastModifiedTime: ' ', $
> Size: 0L }
>
> FrameInformation = { Width: 0L, $
> Height: 0L }
>
> ConditionsFile = { Properties: Properties, $
> Frame_Information: FrameInformation }
>
> SavedFile = { ...etc }
>
>
> FilesAttributes = { ConditionsFile: ConditionsFile, $
> SavedFile: SavedFile }
>
> etc..etc.
>
> although your lines may get a bit long, e.g.
>
> print, FilesAttributes.ConditionsFiles.Properties.LastModifiedTime
>
> ...phew!
>
> I would use structure contructor functions (search for "Automatic
Structure Definition" in
> the IDL help) to do it, so you can automatically define a valid structure
containing all
> the allowed bits. You could also approach this in an OO way (search for
"IDL Object
> Overview"). I'm not an IDL OO feller, so others will have more to say
about that aspect.
>
> All the data stuff would be stored in a pointer array structure element
that you can
> allocate when you determine what the required dimensions are.
>
> Actually it sounds like a neato little project to do this. you'll learn
everything you
> ever wanted to know about IDL structures, and if you're keen, about the OO
capabilities
> also.
>
> paulv
>
> --
> Paul van Delst
> CIMSS @ NOAA/NCEP/EMC
> Ph: (301)763-8000 x7274
> Fax:(301)763-8545
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34196 is a reply to message #34194] |
Thu, 20 February 2003 11:54 |
CelticBlues
Messages: 10
Registered: February 2003
|
Junior Member |
|
|
See below....
"David Fanning" <david@dfanning.com> wrote in message
news:MPG.18bed78a6fdf4fbf989af2@news.frii.com...
> CelticBlues (idluser@celticblues.com) writes:
>
>> The file is a home grown tagged file format containing a type of image
data
>> and additional information about the conditions at the time the image
was
>> taken etc. Here is an example of what a simple file may look like:
>
> Oh, no. I'm not going to be sucked into this. I have my
> own work to do. (Actually, now that I think about it, I
> am a little short of paying customers at the moment...)
> But, no, I'm not going to do this.
Darn... you mean you won't write it for me? ;-) Actually the info you give
below is just what I need... An idea of where to look for the answer. I
don't expect anyone to hand me the complete solution (where's the fun in
that) but a little push in the right direction...
Thanks,
Ed
> And, for heaven's sake, get a book about IDL programming.
> If you are a serious sort of person get Liam Gumley's
> book. If you prefer something with a bit more attitude,
> get my book. But you are unlikely to learn what you
> really want to know by looking at the IDL documentation.
> Either one of these books will parse the essential from
> the overwhelming for you. :-)
Not sure what you mean by "a serious sort of person"... but I like books
that are are somewhat tutorial like, but aren't limited to overly simple
examples. The OpenGL Red Book a good example of the type of book I like to
learn a new topic with....
Later,
Ed
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34197 is a reply to message #34196] |
Thu, 20 February 2003 11:17 |
Rick Towler
Messages: 821
Registered: August 1998
|
Senior Member |
|
|
"CelticBlues" wrote in
> It's a fairly complex ascii file, the parser of which I did not write
> myself. Combine the complexity, with the fact that I didn't write the
> original parser, and I feel more comfortable reusing the code, than
> rewriting it in IDL, which I just started using yesterday. Maybe as I
learn
> more about IDL, I'll write a parser in IDL, but until then....
Everyone seems to be chiming in with their opinions and they are all valid.
Determining where to spend your effort is difficult, especially if you don't
have much experience with IDL and you don't have intimate knowledge of the
file format. You can really help your cause if you post a snippet of your
file since some of the experienced IDL'ers can give you their opinion on the
level of effort needed to parse a record in IDL.
If you decide to go the C++ route I would suggest coding a wrapper function
in C++ to interface your code to IDL as an dynamically loadable module (aka
dlm). I know you don't have much experience with IDL. I am hoping you have
some with C/C++? As a novice C/C++ coder I have written a number of simple
.dlms. If you have experience with C/C++ you should be well on your way.
How simple is it to use the C++ code? Do you just pass the function a file
name and it returns a struct or do you pass it a record as a char array and
it returns the parsed record in a struct? The simpler to use, the simpler
to code into a dlm.
Also, if you go the dlm route I strongly suggest picking up a copy of Ronn
Kling's book "Calling C from IDL" available from his website
www.kilvarock.com.
I could probably work up an example of a dlm which returns a structure to
IDL if you are interested.
-Rick
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34198 is a reply to message #34197] |
Thu, 20 February 2003 11:28 |
Paul Van Delst[1]
Messages: 1157
Registered: April 2002
|
Senior Member |
|
|
CelticBlues wrote:
>
> The file is a home grown tagged file format containing a type of image data
> and additional information about the conditions at the time the image was
> taken etc. Here is an example of what a simple file may look like:
> ---------------------------------------------------------
> [System]
> Name = 1
> [End]
> [Data Sets] // will be at least one data set, could be as many as 3
> DataSet1
> [End]
> [Files Attributes]
> [Conditions File]
> [Properties]
> Name = 97918.10
> CreateTime = Thu Sept 18 1997 at 11:21:36
> LastAccessTime = Mon Jan 01 1601
> LastModifiedTime = Thu Sept 18 1997 at 11:21:36
> Size = 16882000 // bytes
> [End]
> [Frame information]
> Width = 133
> Height = 148
> [End]
> [End] // end [Telemetry File]
>
> [Saved File]
> CreateDate = Tue Feb 11 2003 at 16:08:54
> Width = 133
> Height = 148
> FrameNumber = 1
> DataUnits = meters
> Comment =
> CoordinateSystem = UTM // Latlong or UTM
> [End]
>
> [End] // end [Files Attributes]
>
> [DataBlock:DataSet1]
> R1P1 R1P2...R1P148
> R2P1 R2P2...R2P148
> .
> .
> .
> R133P1 R133P2...R133P148
> [End]
>
> ---------------------------------------------------------
>
> R1P1 = Row 1 Pixel 1, R2P1 = Row 2, Pixel 2 etc.
>
> One additional bit of info... the datablock may not actually have CR/NL
> after each row, i.e. it may be
>
> [DataBlock:DataSet1]
> R1P1 R1P2...R1P148 R2P1 R2P2...R2P148 ... R133P1 R133P2...R133P148
> [End]
>
> Ok, assume that I want to rewrite the parser in IDL... From what I
> understand it is simple enough to read an ascii table from a file, but what
> about reading this type of format?
The reading isn't the problem - it's the sorting of the read-in bits. :o)
But, it doesn't look too difficult. Given what you know about the allowable syntax in your
ASCII file, you can create the structures up front, e.g.
Properties = { Name: 0.0, $
CreateTime: ' ', $
LastAccessTime: ' ', $
LastModifiedTime: ' ', $
Size: 0L }
FrameInformation = { Width: 0L, $
Height: 0L }
ConditionsFile = { Properties: Properties, $
Frame_Information: FrameInformation }
SavedFile = { ...etc }
FilesAttributes = { ConditionsFile: ConditionsFile, $
SavedFile: SavedFile }
etc..etc.
although your lines may get a bit long, e.g.
print, FilesAttributes.ConditionsFiles.Properties.LastModifiedTime
...phew!
I would use structure contructor functions (search for "Automatic Structure Definition" in
the IDL help) to do it, so you can automatically define a valid structure containing all
the allowed bits. You could also approach this in an OO way (search for "IDL Object
Overview"). I'm not an IDL OO feller, so others will have more to say about that aspect.
All the data stuff would be stored in a pointer array structure element that you can
allocate when you determine what the required dimensions are.
Actually it sounds like a neato little project to do this. you'll learn everything you
ever wanted to know about IDL structures, and if you're keen, about the OO capabilities
also.
paulv
--
Paul van Delst
CIMSS @ NOAA/NCEP/EMC
Ph: (301)763-8000 x7274
Fax:(301)763-8545
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34199 is a reply to message #34197] |
Thu, 20 February 2003 11:16 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
CelticBlues (idluser@celticblues.com) writes:
> The file is a home grown tagged file format containing a type of image data
> and additional information about the conditions at the time the image was
> taken etc. Here is an example of what a simple file may look like:
Oh, no. I'm not going to be sucked into this. I have my
own work to do. (Actually, now that I think about it, I
am a little short of paying customers at the moment...)
But, no, I'm not going to do this.
This is not a hard file format to figure out.
Learn how to read a line of the file at a time:
line = ""
OPENR, lun, 'myfile.dat', /Get_Lun
ReadF, lun, line
Learn how to parse that line. I'd pay particular
attention to STRPOS and STRMID, for starters.
I'd learn what FIX, LONG, and FLOAT can do.
I don't know what kind of data that is in the data
block, but presumably those letters mean *something*.
I'd learn how to decode that.
There is a day's worth of material here, for sure.
But this is a LOT easier than you are probably
thinking right now. Try it out by doing things
at the IDL command line. IDL is fabulous at what
I call "learning by noodling around". You have
a tremendous advantage of being able to see and make
sense of what is in the data file. We are not all
so lucky.
And, for heaven's sake, get a book about IDL programming.
If you are a serious sort of person get Liam Gumley's
book. If you prefer something with a bit more attitude,
get my book. But you are unlikely to learn what you
really want to know by looking at the IDL documentation.
Either one of these books will parse the essential from
the overwhelming for you. :-)
Good luck. I think you are going to find out you
LOVE IDL (well, eventually). :-)
Cheers,
David
--
David W. Fanning, Ph.D.
Fanning Software Consulting, Inc.
Phone: 970-221-0438, E-mail: david@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34200 is a reply to message #34199] |
Thu, 20 February 2003 11:11 |
CelticBlues
Messages: 10
Registered: February 2003
|
Junior Member |
|
|
For me, it would be ideal, if I could do everything in IDL...but I don't
work in an ideal world... So I think your suggestion of an intermediate file
written by the original parser is probably the way to go, although I still
have the IDL parser on the back burner. That will give me a fall back, if I
can't get an ascii reader written in IDL by dinner time (as David Fanning
suggests :-)....
Thanks to everyone for the comments,
Ed
"Paul van Delst" <paul.vandelst@noaa.gov> wrote in message
news:3E552256.3BF2399F@noaa.gov...
> CelticBlues wrote:
>>
>> It's a fairly complex ascii file, the parser of which I did not write
>> myself. Combine the complexity, with the fact that I didn't write the
>> original parser, and I feel more comfortable reusing the code, than
>> rewriting it in IDL, which I just started using yesterday. Maybe as I
learn
>> more about IDL, I'll write a parser in IDL, but until then....
>
> I'm not being flippant when I say this may be a good opportunity to learn
IDL - and
> understand more about your data file/structure. Assuming you have the
luxury of time.
>
> Otherwise, why not use the C++ parser to read the file and write some
simple little C++
> program to output the structure in some really simple format that would
take "5 minutes"
> to write a parser for to read into IDL? (Here I'm assuming all you want to
use IDL for in
> this case is plot the data)
>
> paulv
>
>
>>
>> Thanks for the info,
>>
>> Later,
>> Ed
>>
>> "David Fanning" <david@dfanning.com> wrote in message
>> news:MPG.18bec429c19cf041989af0@news.frii.com...
>>> James Kuyper (kuyper@saicmodis.com) writes:
>>>
>>>> That depends entirely on how complicated the syntax that it parses
is.
>>>> If he's worried about the recoding effort, I doubt that the parsing
is
>>>> simple enough to be a 5 minute job in IDL. With all due respect to
IDL,
>>>> most things that can be parsed correctly by a 5-minute IDL program
have
>>>> a fairly simple syntax.
>>>
>>> Perhaps I exaggerated a little. (It's been known to happen.)
>>> But I stand by my "orders of magnitude" comment. :-)
>>>
>>> Cheers,
>>>
>>> David
>>>
>>> --
>>> David W. Fanning, Ph.D.
>>> Fanning Software Consulting, Inc.
>>> Phone: 970-221-0438, E-mail: david@dfanning.com
>>> Coyote's Guide to IDL Programming: http://www.dfanning.com/
>>> Toll-Free IDL Book Orders: 1-888-461-0155
>
> --
> Paul van Delst
> CIMSS @ NOAA/NCEP/EMC
> Ph: (301)763-8000 x7274
> Fax:(301)763-8545
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34201 is a reply to message #34200] |
Thu, 20 February 2003 10:51 |
CelticBlues
Messages: 10
Registered: February 2003
|
Junior Member |
|
|
The file is a home grown tagged file format containing a type of image data
and additional information about the conditions at the time the image was
taken etc. Here is an example of what a simple file may look like:
---------------------------------------------------------
[System]
Name = 1
[End]
[Data Sets] // will be at least one data set, could be as many as 3
DataSet1
[End]
[Files Attributes]
[Conditions File]
[Properties]
Name = 97918.10
CreateTime = Thu Sept 18 1997 at 11:21:36
LastAccessTime = Mon Jan 01 1601
LastModifiedTime = Thu Sept 18 1997 at 11:21:36
Size = 16882000 // bytes
[End]
[Frame information]
Width = 133
Height = 148
[End]
[End] // end [Telemetry File]
[Saved File]
CreateDate = Tue Feb 11 2003 at 16:08:54
Width = 133
Height = 148
FrameNumber = 1
DataUnits = meters
Comment =
CoordinateSystem = UTM // Latlong or UTM
[End]
[End] // end [Files Attributes]
[DataBlock:DataSet1]
R1P1 R1P2...R1P148
R2P1 R2P2...R2P148
.
.
.
R133P1 R133P2...R133P148
[End]
---------------------------------------------------------
R1P1 = Row 1 Pixel 1, R2P1 = Row 2, Pixel 2 etc.
One additional bit of info... the datablock may not actually have CR/NL
after each row, i.e. it may be
[DataBlock:DataSet1]
R1P1 R1P2...R1P148 R2P1 R2P2...R2P148 ... R133P1 R133P2...R133P148
[End]
Ok, assume that I want to rewrite the parser in IDL... From what I
understand it is simple enough to read an ascii table from a file, but what
about reading this type of format?
Ed
"David Fanning" <david@dfanning.com> wrote in message
news:MPG.18beb4cde3551d28989aef@news.frii.com...
> CelticBlues (idluser@celticblues.com) writes:
>
>> I have an ascii data file that I need to read. I already have C++ code
that
>> will parse the file and store it in a structure. Is there a way to
reuse
>> this C++ code in IDL so that I don't have to recode the parse routines?
>
> Yes, there is a way to reuse this code, but learning
> the methods to do it will far surpass (by many orders of
> magnitude) the time it will take (say 5 minutes) to
> learn to parse an ASCII file in IDL. :-)
>
> See Call_External, Make_DLL, LinkImage, etc.
>
> Cheers,
>
> David
> --
> David W. Fanning, Ph.D.
> Fanning Software Consulting, Inc.
> Phone: 970-221-0438, E-mail: david@dfanning.com
> Coyote's Guide to IDL Programming: http://www.dfanning.com/
> Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34202 is a reply to message #34201] |
Thu, 20 February 2003 10:52 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
CelticBlues (idluser@celticblues.com) writes:
> It's a fairly complex ascii file, the parser of which I did not write
> myself. Combine the complexity, with the fact that I didn't write the
> original parser, and I feel more comfortable reusing the code, than
> rewriting it in IDL, which I just started using yesterday. Maybe as I learn
> more about IDL, I'll write a parser in IDL, but until then....
My only point is that learning how to link C++ programs
into IDL on your second day on the job, is like learning
C++ by starting with how the compiler works. It may be
useful (and even interesting), but at the end of the
day you aren't going to be much of anywhere.
But if you spend a day with IDL futzing around trying
to read an ASCII file, you might even have a file
reader by the time the dinner bell goes off. :-)
Cheers,
David
--
David W. Fanning, Ph.D.
Fanning Software Consulting, Inc.
Phone: 970-221-0438, E-mail: david@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34203 is a reply to message #34201] |
Thu, 20 February 2003 10:45 |
Paul Van Delst[1]
Messages: 1157
Registered: April 2002
|
Senior Member |
|
|
CelticBlues wrote:
>
> It's a fairly complex ascii file, the parser of which I did not write
> myself. Combine the complexity, with the fact that I didn't write the
> original parser, and I feel more comfortable reusing the code, than
> rewriting it in IDL, which I just started using yesterday. Maybe as I learn
> more about IDL, I'll write a parser in IDL, but until then....
I'm not being flippant when I say this may be a good opportunity to learn IDL - and
understand more about your data file/structure. Assuming you have the luxury of time.
Otherwise, why not use the C++ parser to read the file and write some simple little C++
program to output the structure in some really simple format that would take "5 minutes"
to write a parser for to read into IDL? (Here I'm assuming all you want to use IDL for in
this case is plot the data)
paulv
>
> Thanks for the info,
>
> Later,
> Ed
>
> "David Fanning" <david@dfanning.com> wrote in message
> news:MPG.18bec429c19cf041989af0@news.frii.com...
>> James Kuyper (kuyper@saicmodis.com) writes:
>>
>>> That depends entirely on how complicated the syntax that it parses is.
>>> If he's worried about the recoding effort, I doubt that the parsing is
>>> simple enough to be a 5 minute job in IDL. With all due respect to IDL,
>>> most things that can be parsed correctly by a 5-minute IDL program have
>>> a fairly simple syntax.
>>
>> Perhaps I exaggerated a little. (It's been known to happen.)
>> But I stand by my "orders of magnitude" comment. :-)
>>
>> Cheers,
>>
>> David
>>
>> --
>> David W. Fanning, Ph.D.
>> Fanning Software Consulting, Inc.
>> Phone: 970-221-0438, E-mail: david@dfanning.com
>> Coyote's Guide to IDL Programming: http://www.dfanning.com/
>> Toll-Free IDL Book Orders: 1-888-461-0155
--
Paul van Delst
CIMSS @ NOAA/NCEP/EMC
Ph: (301)763-8000 x7274
Fax:(301)763-8545
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34205 is a reply to message #34203] |
Thu, 20 February 2003 10:12 |
CelticBlues
Messages: 10
Registered: February 2003
|
Junior Member |
|
|
It's a fairly complex ascii file, the parser of which I did not write
myself. Combine the complexity, with the fact that I didn't write the
original parser, and I feel more comfortable reusing the code, than
rewriting it in IDL, which I just started using yesterday. Maybe as I learn
more about IDL, I'll write a parser in IDL, but until then....
Thanks for the info,
Later,
Ed
"David Fanning" <david@dfanning.com> wrote in message
news:MPG.18bec429c19cf041989af0@news.frii.com...
> James Kuyper (kuyper@saicmodis.com) writes:
>
>> That depends entirely on how complicated the syntax that it parses is.
>> If he's worried about the recoding effort, I doubt that the parsing is
>> simple enough to be a 5 minute job in IDL. With all due respect to IDL,
>> most things that can be parsed correctly by a 5-minute IDL program have
>> a fairly simple syntax.
>
> Perhaps I exaggerated a little. (It's been known to happen.)
> But I stand by my "orders of magnitude" comment. :-)
>
> Cheers,
>
> David
>
> --
> David W. Fanning, Ph.D.
> Fanning Software Consulting, Inc.
> Phone: 970-221-0438, E-mail: david@dfanning.com
> Coyote's Guide to IDL Programming: http://www.dfanning.com/
> Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34206 is a reply to message #34205] |
Thu, 20 February 2003 09:54 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
James Kuyper (kuyper@saicmodis.com) writes:
> That depends entirely on how complicated the syntax that it parses is.
> If he's worried about the recoding effort, I doubt that the parsing is
> simple enough to be a 5 minute job in IDL. With all due respect to IDL,
> most things that can be parsed correctly by a 5-minute IDL program have
> a fairly simple syntax.
Perhaps I exaggerated a little. (It's been known to happen.)
But I stand by my "orders of magnitude" comment. :-)
Cheers,
David
--
David W. Fanning, Ph.D.
Fanning Software Consulting, Inc.
Phone: 970-221-0438, E-mail: david@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34207 is a reply to message #34206] |
Thu, 20 February 2003 09:08 |
James Kuyper
Messages: 425
Registered: March 2000
|
Senior Member |
|
|
David Fanning wrote:
>
> CelticBlues (idluser@celticblues.com) writes:
>
>> I have an ascii data file that I need to read. I already have C++ code that
>> will parse the file and store it in a structure. Is there a way to reuse
>> this C++ code in IDL so that I don't have to recode the parse routines?
>
> Yes, there is a way to reuse this code, but learning
> the methods to do it will far surpass (by many orders of
> magnitude) the time it will take (say 5 minutes) to
> learn to parse an ASCII file in IDL. :-)
That depends entirely on how complicated the syntax that it parses is.
If he's worried about the recoding effort, I doubt that the parsing is
simple enough to be a 5 minute job in IDL. With all due respect to IDL,
most things that can be parsed correctly by a 5-minute IDL program have
a fairly simple syntax.
|
|
|
|
| Re: Is there a way to use C++ code in IDL [message #34208 is a reply to message #34207] |
Thu, 20 February 2003 08:48 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
CelticBlues (idluser@celticblues.com) writes:
> I have an ascii data file that I need to read. I already have C++ code that
> will parse the file and store it in a structure. Is there a way to reuse
> this C++ code in IDL so that I don't have to recode the parse routines?
Yes, there is a way to reuse this code, but learning
the methods to do it will far surpass (by many orders of
magnitude) the time it will take (say 5 minutes) to
learn to parse an ASCII file in IDL. :-)
See Call_External, Make_DLL, LinkImage, etc.
Cheers,
David
--
David W. Fanning, Ph.D.
Fanning Software Consulting, Inc.
Phone: 970-221-0438, E-mail: david@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
