How to change settings for just a few readings?
I'm looking at how to convert from .svx format to .th format. Mostly working, except I can't see how to handle this cleanly (the situation modelled here is that the tape is too short, and for one leg is was measured from a different point - in the example this was distilled from because 2-3 was a plumb and we needed more tape to tie a crab on): *begin test *calibrate tape 1 1 2 10 0 0 *begin *calibrate tape 2 2 3 10 0 0 *end ; *calibrate 1 is now be back in effect 3 4 10 0 0 *end test The nearest I can get is: encoding iso8859-1 survey test -title "test" centreline calibrate tape 1 1 2 10 0 0 endcentreline centreline calibrate tape 2 2 3 10 0 0 endcentreline centreline 3 4 10 0 0 endcentreline endsurvey However this doesn't restore the "calibrate tape 1" of course (the survey length is 27m). I can't omit the survey name for a survey command, or pass an empty name. And I can't nest centreline commands. The only solution I can see is to track which settings are in effect and reset them all after the *end. That's just about OK for automated conversion if nobody's going to want to edit the result, but it must be a right headache for anyone who has tried to hand convert a dataset so I'm wondering if there's some way to express this situation. Cheers, Olly
Olly Betts wrote:
The only solution I can see is to track which settings are in effect and reset them all after the *end. That's just about OK for automated conversion if nobody's going to want to edit the result, but it must be a right headache for anyone who has tried to hand convert a dataset so I'm wondering if there's some way to express this situation.
No, there is no such possibility. You have to repeat calibration every time it changes. I would rewrite your example this way: encoding iso8859-1 survey test -title "test" centreline calibrate tape 1 1 2 10 0 0 calibrate tape 2 2 3 10 0 0 calibrate tape 1 3 4 10 0 0 endcentreline endsurvey I'm just wondering, whether these situations are so common... S.
On Tue, Sep 28, 2004 at 12:29:51PM +0200, Stacho Mudrak wrote:
Olly Betts wrote:
The only solution I can see is to track which settings are in effect and reset them all after the *end. That's just about OK for automated conversion if nobody's going to want to edit the result, but it must be a right headache for anyone who has tried to hand convert a dataset so I'm wondering if there's some way to express this situation.
No, there is no such possibility. You have to repeat calibration every time it changes. I would rewrite your example this way:
encoding iso8859-1 survey test -title "test" centreline calibrate tape 1 1 2 10 0 0 calibrate tape 2 2 3 10 0 0 calibrate tape 1 3 4 10 0 0 endcentreline endsurvey
Ick.
I'm just wondering, whether these situations are so common...
It happens 8 times in CUCC's Austria dataset. Cheers, Olly
Stacho Mudrak wrote:
No, there is no such possibility. You have to repeat calibration every time it changes. I would rewrite your example this way:
encoding iso8859-1 survey test -title "test" centreline calibrate tape 1 1 2 10 0 0 calibrate tape 2 2 3 10 0 0 calibrate tape 1 3 4 10 0 0 endcentreline endsurvey
I'm just wondering, whether these situations are so common...
I'd say common enough on fairly large survey projects, especially those under exploration/expedition conditions where the surveying is done in a hurry and the chances of going back and correcting iffy readings are low. I quite often make use of changing the SD for particular legs where it's known that the survey wasn't done to the general standard. In general the SD is defined in a separate file which I *include in the survex data. Having to undo that separation when using therion, and repeat the definition of the SD, is a shame, and makes converting data harder (as I first attempt I've usually just ignored changes in SD and put a comment in the .th file). Yours, John
There is no problem to add somethig like group-endgroup pair inside centerline, if it will really help you (or I can use just begin-end pair if you wish). Then it could be rewritten this way: centreline calibrate tape 1 1 2 10 0 0 group calibrate tape 2 2 3 10 0 0 endgroup 3 4 10 0 0 endcentreline Or do you see some other trivial solution of this problem? Regards, S. John Pybus wrote:
Stacho Mudrak wrote:
No, there is no such possibility. You have to repeat calibration every time it changes. I would rewrite your example this way:
encoding iso8859-1 survey test -title "test" centreline calibrate tape 1 1 2 10 0 0 calibrate tape 2 2 3 10 0 0 calibrate tape 1 3 4 10 0 0 endcentreline endsurvey
I'm just wondering, whether these situations are so common...
I'd say common enough on fairly large survey projects, especially those under exploration/expedition conditions where the surveying is done in a hurry and the chances of going back and correcting iffy readings are low.
I quite often make use of changing the SD for particular legs where it's known that the survey wasn't done to the general standard. In general the SD is defined in a separate file which I *include in the survex data. Having to undo that separation when using therion, and repeat the definition of the SD, is a shame, and makes converting data harder (as I first attempt I've usually just ignored changes in SD and put a comment in the .th file).
Yours,
John
Stacho Mudrak wrote:
There is no problem to add somethig like group-endgroup pair inside centerline, if it will really help you (or I can use just begin-end pair if you wish).
Then it could be rewritten this way:
centreline calibrate tape 1 1 2 10 0 0 group calibrate tape 2 2 3 10 0 0 endgroup 3 4 10 0 0 endcentreline
Or do you see some other trivial solution of this problem?
That seems like the most natural solution; it directly mirrors an unlabeled begin-end pair in survex data. I see no need for the syntax to actually be begin/end, and group/endgroup sounds ideal. Thanks, it would tidy up one of the few edge cases in mapping survex data to therion centrelines. John
OK, I'll put it on the top of TODO list - at least now it seems to be like a very simple thing to implement. Regards, S. John Pybus wrote:
Stacho Mudrak wrote:
There is no problem to add somethig like group-endgroup pair inside centerline, if it will really help you (or I can use just begin-end pair if you wish).
Then it could be rewritten this way:
centreline calibrate tape 1 1 2 10 0 0 group calibrate tape 2 2 3 10 0 0 endgroup 3 4 10 0 0 endcentreline
Or do you see some other trivial solution of this problem?
That seems like the most natural solution; it directly mirrors an unlabeled begin-end pair in survex data. I see no need for the syntax to actually be begin/end, and group/endgroup sounds ideal. Thanks, it would tidy up one of the few edge cases in mapping survex data to therion centrelines.
John
The attached patch should enable group-endgroup inside centerline command. I have not tested it extensively, but you should be able to change everything (data order, calibration, sd, units etc.) inside. Regards, S. John Pybus wrote:
Stacho Mudrak wrote:
There is no problem to add somethig like group-endgroup pair inside centerline, if it will really help you (or I can use just begin-end pair if you wish).
Then it could be rewritten this way:
centreline calibrate tape 1 1 2 10 0 0 group calibrate tape 2 2 3 10 0 0 endgroup 3 4 10 0 0 endcentreline
Or do you see some other trivial solution of this problem?
That seems like the most natural solution; it directly mirrors an unlabeled begin-end pair in survex data. I see no need for the syntax to actually be begin/end, and group/endgroup sounds ideal. Thanks, it would tidy up one of the few edge cases in mapping survex data to therion centrelines.
John
participants (3)
-
John Pybus -
Olly Betts -
Stacho Mudrak