Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/sas/6.0.0/doc/odfingest/ChangeLog Дата изменения: Fri Dec 12 20:15:27 2003 Дата индексирования: Sat Dec 22 09:07:38 2007 Кодировка: Поисковые слова: http astrokuban.info astrokuban

ChangeLog for odfingest
=======================

Version 3.23.1 - 2003-12-12 (JSH)
------------
+ Fix Makefiles in New odfingest to fix bizarre run-time error caused
by order of libraries in link phase of dynamic build. No, don't ask.

Version 3.23 - 2003-11-12 (JSH)
------------
+ (TestTimeRows.cc) Catch end() used in first test, don't print
end() results in second test.
+ (HkparameterTest.cc) Modify duration to process to match reality

Version 3.22 - 2003-08-29 (GV)
------------
+ (TimeRows.cc) Get rid of a message that was being printed 100's
of times without being informative.
+ (new) test harnesses: adjust a couple of constructors that had
the wrong signature after Hkparameter had been changed.

Version 3.21 - 2003-08-26 (GV)
------------
+ Several fixes to nodfingest, test harnesses extended to cover
the problems that showed up in testing.

Version 3.20 - 2003-06-10 (GV)
------------
+ ( Hkparameter_Descriptor.h) Parameter K1211 was incorrectly
assigned to MOS1 instead of MOS2.

Version 3.19 - 2003-05-13 (GV)
------------
+ Add CMCORR parameters for PN (F1525/1625/1725/1825).

Version 3.18 - 2003-05-07 (GV)
------------
+ Add parameter MEAN_TEMPERATURE (E/K 1311) for MOS.

Version 3.17 - 2003-05-06 (GV)
------------
+ Add next-generation odfingest, temporarily called nodfingest.

Version 3.16 - 2003-02-21 (GV)
------------
+ (odfingest.cc) Make the name of the summary file available to
calling meta-tasks.

Version 3.15 - 2002-10-14 (GV)
------------
+ (src/*.cc) remove a few remaining dts::ends that were apparently
confusing selectlib (SOC-SPR-2350).

Version 3.14 - 2002-09-17 (GV)
------------
+ (testtot.cc) include so that it compiles on gcc-2.96.

Version 3.13 - 2002-09-16 (GV)
------------
+ Misc. source/header adjustments for gcc-3.1

Version 3.12 - 2002-07-09 (GV)
------------
+ Restore DEPEND.

Version 3.11 - 2002-07-08 (GV)
------------
+ DEPEND updated.

Version 3.10 - 2002-05-17 (GV)
------------
+ updated for param 2.0

Version 3.9 - 2002-05-17 (GV)
-----------
+ (OdfObject.cc) make the messages about missing CONFIGURATIOn or
PROPOSAL record more explicit.

Version 3.8 - 2002-04-10 (GV)
-----------
+ (OdfObject.cc::proposal_summary_rec) Change how the PROPOSAL
record is read from the original SUM.ASC file. This should make it
more robust against those cases when the line of text containing
the word PROPOSAL contains characters other than blank. SOC-SPR-2286.
+ A missing CONFIGURATION record in the SUM.ASC is no longer a
warning, only a message.

Version 3.7.5 - 2002-03-18 (GV)
-------------
+ Test harness against DCP 141 updated (odfconf). Relevant section
was actually commented out.

Version 3.7.4 - 2002-03-05 (GV)
-------------
+ (IppvImplementation.cc) When no translation from a value to
string is found, IppvCommon::print writes UNDEFINED instead of
NoTranslation in the summary file.
+ (OdfObject.cc:proposal_summary_rec()) copy the SUM.ASC into the
SUM.SAS from the PROPOSAL section or, if present, the new
CONFIGURATION section (SOC RCR 638, ODF ICD DCP 141).

Version 3.7.3 - 2002-02-27 (UL)
-------------
+ fixed problem with incorrect Dal usage in test/testtot.cc
(caused test failure under Linux)

Version 3.7.2 - 2002-02-14 (GV)
-------------
+ (IppvImplementation.cc) setValue() tolerates one HK record with
a different value than expected. In a few cases in fact (for
instance 0102/0099020301, PNS012) the value for F1118 (FILTER)
changes just for 8s to CLOSE (it was MEDIUM). Consider this a
glitch in tm, and carry on.
+ (OdfObject.cc) DLI and DII files are now given an arbitrary
duration of 1s instead of 4m as before. Giving them 4m may
make exposures with both science and DLI/DII file too long -> HK
extraction fails (see: 0101/0124710301, PNU002).

Version 3.7.1 - 2002-01-28 (GV)
-------------
+ (OdfObject.cc) change tpad to 240s (was 90). There are two
consequences:
. the default arbitrary exposure duration for some type of
exposures becomes 480s (it is coded as 2*tpad).
. read the HK values from Tstart(exposure) + 240s to
Tstop(exposure) - 240s

Version 3.7 - 2002-01-24 (GV)
-----------
+ (IppvExtractionImplementation.cc) Add GAIN_CCD for MOS. This
required also the addition of EmosGainTranslator.

Version 3.6 - 2002-01-17 (GV)
-----------
+ ((OdfObject.cc) handle the case when DLI + science mode data are
present in the same exposure, by giving priority to the science
data.

Version 3.5 - 2002-01-14 (GV)
-----------
+ (IppvExtractionImplementation.c) PN CAMEX_GAIN for CCD4 should
have used F1234 instead of F1243.

Version 3.4 - 2002-01-07 (GV)
-----------
+ (OdfObject.cc::scan) Modify slightly the algorithm used when
DATE-END is not a valid time string. A dataset with an invalid
DATE-END used to be assigned an arbitrary duration of 180s. But
there are cases (mostly ODFs from the real time chain) that can
have an invalid DATE-END and contain good data. Now the algorithm
will, if possible, set DATE-END to be the time of the last record
in the main HK dataset.
+ (IppvExtractionImplementation.cc) Add CAMEX_GAIN_CCDx for PN
(SOC-SPR-2154). Doc updated.

Version 3.3 - 2001-11-15 (GV)
-----------
+ re-upload of 3.2.8 with new version number for reltrack.

Version 3.2.9 - 2001-11-15 (GV)
-------------
+ (src/OdfObject.cc) undo the i/f changes on OdfSet (3.2.4) so
that it can work with oal as in SAS 5.2.0. NOTE: DEPEND cannot be
updated meaningfully because we have oal-3.68.6 in SAS 5.2.0 and
the interface change happened at oal-3.68.8 (as of today in
release track).

Version 3.2.8 - 2001-11-06 (GV)
------------
+ (src/OdfObject.cc) Now also cope correctly with the case of TIE
+ NOI/ODI, neglected during the previous updates.

Version 3.2.7 - 2001-11-02 (GV)
-------------
+ (src/OdfObject.cc) cope 'correctly' with the case of both NOI and
IME data files in the same PN exposure.
+ SLEW identification: 9RRRR#####, but not 99999#####, as these
are unscheduled observations.

Version 3.2.6 - 2001-10-30 (GV)
-------------
+ (src/OdfObject.cc:scan) impose fixed duration for RGS DII
exposures, as DATE-END is not reliable.
+ (src/OdfObject.cc::scan) determine latest possible DATE-END from
instrument HK. If DATE-END from any dataset appears to be later
than what reported in HK, use the value from the HK instead.
+ (src/OdfObject.cc:scan) catch STime errors when evaluating
DATE-END. If an error occurs, set DATE-END to be start + 180s.
+ (OdfObject::observation_rec) write SLEW instead of OBSERVATION
in the summary file if the observation identifier starts with a 9.
+ (OdfObject::configuration_sub_record) do not use any of the OM
auxiliary datasets in the determination of the instrument mode.
+ (OdfObject.cc) cope 'correctly' with the case of both ODI and
IME data files in the same PN exposure.

Version 3.2.5 - 2001-10-30 (GV)
-------------
+ (calmode.cc) here too I should have adjusted the i/f to OdfSet.
+ (OdfObject.cc) remove condition whereby the cal was asked to
identify the mode only if the dataset was science-related. This to
increase the amount of information in the summary file (and hence
in the ODF d/b and odfbrowser).
+ (src/OdfObject.cc:observation_rec) if there are no science files
in the ODF, then try to determine the observation duration from
the HK files. SOC-SPR-2094. This also required a change in scan(),
where the ODFcan now also be identified from the names of the HK
files.
+ (src/IppvTranslator.cc:EpnFilterWheelTranslator) All CAL
positions are identified as Xmm::FILTER_NONE instead of the
arbitrary "CalibrationPosition" that was breaking the
OAL. SOC-SPR-2114.

Version 3.2.4 - 2001-10-26 (GV)
-------------
+ OdfObject.cc: adjust how OdfSet is created, as the OAL in the
reltrack decided to change this interface (!).
+ DEPEND updated accordingly.

Version 3.2.3 - 2001-10-09 (GV)
-------------
+ Fix doco: parameter oalcheck defaults to no (SOC-SPR-2193).

Version 3.2.2 - 2001-10-05 (GV)
-------------
+ (OdfObject.cc) when determining the OM instrument mode, ask to
window 0 instead of window 1, otherwise the mode determination
fails when there is only one OM window (SOC-SPR-2195).

Version 3.2.1 - 2001-07-18 (UL)
-------------
+ fixed problem with too fragile test test/testtot.cc (see comments)

Version 3.2 - 2001-07-12 (GV)
-----------
+(src/TimeOrderedTble.cc) When the table is constructed time
strings that cannot be translated to a valid time are
ignored. That means that a TOT could have less rows than the table
it was constructed from. A new method (validTimes) introduced to
report on the number of good time entries. When a time is accessed
through time_of_row(k), the k is mean to refer to the orginal
table instead of the TOT. This fixes SOC-SPR-2095, in relation
to ODF 0061_0122700201.
+ (src/OdfObject.cc) during the calculation of the OM exposure
start/stop make sure to check that the TIME column in the NPHK
file is valid (SOC-SPR-2095).
+ (src/IppvImplementation.cc::setValue) if the value of a
parameter changes during the exposure and that parameter is
supposed to be constant, write the record number and the table
name in the warning message, to make visual checking easier.

Version 3.1.1 - 2001-06-08 (GV)
-------------
+ (src/odfingest.cc) Change the manner of setting the target ODF
directory. If the directory is given on the command line, use that
value, else ask the OAL (that'll read it through SAS_ODF or
--odf). This prevents the OAL from being activated when it is not
the user intention (SOC-SPR-2088). Also check that writing to the
output file completes correctly (SOC-SPR-2089).
+ (doc/odingest_description.tex) updated with the new error
condition.
+ (src/OdfObject.cc) use 'window' and not CCD when referring to OM
(SOC-SPR-1969).


Version 3.1 - 2001-05-18 (GV)
-----------
+ (src/IppvTranslator.cc) EpnFilterWheelTranslator maps HK
names to CAL filter names.
+ (src/IppvExtractorImplementation.cc) F1118 is now used to
determine the filter. Introduce FILTER_POTENTIOMETER to monitor
what the old algorithm does.
+ (src/odfingest_description.tex) updated.

Version 3.0 - 2001-05-08 (GV)
-----------
+ (src/IppvExtractorImplementation.cc) FILTER_WHEEL_POSITION is
now constrained with "F1120 == ACTIVE && F1119 == STOPPED".
+ (src/IppvExtractorImplementation.cc) MOS CLOCK_RESET_TIME_*
should have been treated as periodic parameters instead of
non-periodic (SOC-SPR-2035).
+ (doc/odfingest_description.tex) regenerate the documentation of
the parameters to take into account PN filter changes.

Version 2.2.1 - 2001-05-07 (GV)
-------------
+ (doc/odfingest_description.tex) use appropriate macros to refer
to the cal and the oal.
+ (tools/calmode.cc) change the output format.

Version 2.2 - 2001-04-04 (UL)
-----------
+ brought in line with caloalutils-1.47
(Xmm::ATTITUDE -> Xmm::ATTITUDE_HISTORY)
+ updated old-style include statements

Version 2.1 - 2001-04-02 (GV)
-----------
+ (src/IppvTranslator.cc) EpnFilterWheelTranslator now recognizes
also the string 'NO STOP POS.' (final `.') after somebody decided
to fix the puntuation of the ODF HK (SOC-SPR-1974).
+ (src/IppvExtractorImplementation.cc) fix typo in error message.

Version 2.0 - 2001-03-19 (GV)
-----------
+ (src/odfingest.cc) by default use the ODF pointed to by
SAS_ODF. This is in SAS-SCR-0038.
+ (config/odfingest.par) introduce --withodfdir (default = no).
+ (config/odfingest.lyt) use it.
+ (src/OdfObject.cc) allow OM to have IMI + FAE combinations
(SOC-SPR-1940).

Version 1.0.9 - 2001-02-05 (GV)
-------------
+ SOC-SPR-1907:
- (src/IppvExtractorImplementation.cc) in function hkname change
the check d.size !=1 to d.size > 1. That is, it is ok to have 0
or 1 HK files, but it is not ok to have more than 1.

Version 1.0.8 - 2001-01-29 (GV)
-------------
+ SOC-SPR-1907:
- (src/IppvExtractorImplementation.cc) check if required HK file
exists. If not issue a warning, set all HK parameters to
unknown, and continue.
+ SOC-SPR-1908:
- (src/IppvExtractor*): add _label data member. In OdfObject.cc
the label is set to instrument/exposure, so that warning
messages from the HK extraction process can be made clearer.
- (src/IppvImplementation.cc) HK file and table name,
sought time entry, and available time range are given in the
warning messages.
+ SOC-SPR-1909:
- (src/IppvImplementation.*) warning about new parameter value
changed into an exception throw.
- (src/Ippv.h) add exception ParameterChange.
+ SOC-SPR-1910:
- (src/OdfObject.*) introduce _tpad;
- (src/OdfObject.*) introduce _no_ippv_modes, a list of data
modes that are not considered when extracting IPPVs. For
instance: Discarded Lines Images.
+ (doc/odfingest_description.tex) updates required by the changes
above.
+ SOC-SPR-1915:
- (src/OdfObject.cc) use cmodes.clear() and not cmodes.empty()
(same for dmodes. Former erases all elements, latter checks if
there are any elements. This prevented odfingest from correctly
writing MODE = UNDEFINED when data files of more than one type
where found in one exposure.

Version 1.0.7 - 2000-12-15 (GV)
-------------
+ (src/OdfObject.cc) remove assertions (that are triggered by bad
ODFs more often then they should) and replace them with
errors. Same result, but one gets an error message now. SOC-SPR-1864.
+ (src/OdfObject.cc) introduce member _valid, to indicate whether
a directory scan returned a non empty list of ODF constituents.
+ (src/odfingest.cc) use _valid to prevent segfault and unfriendly
message (SOC-SPR-1855).
+ (doc/odfingest_description.tex) update list of warnings.


Version 1.0.6 - 2000-11-16 (GV)
-------------
+ (src/IppvTranslator.cc) change position of calthin1 filter for PN
fromm 113.85 to 110.65, where it should have always been!

Version 1.0.5 - 2000-11-13 (GV)
-------------
+ (src/OdfObject.cc) make DLI duration 60s.
+ (src/OdfObject.cc) add/subtract 90s (instead of 60) to
DAT-OBS/END when considering the exposure time range over which
HK parameters are to be extracted (SOC-SPR-1833).
+ (src/OdfObject.cc) calculate observation duration more
accurately (SOC-SPR-1831).

Version 1.0.4 - 2000-11-09 (GV)
-------------
+ (odfingest.lyt) filled in.

Version 1.0.3 - 2000-11-08 (GV)
-------------
+ (src/OdfObject.cc): when determining the observation duration,
treat the PN discarded lines data sets in a different manner,
assigning and arbitray duration of 120s to each of them. This to
prevent making the observation unrealistically long (the discarded
lines data sets have the wrong DATE-END, often several hours later
than the actual observation end time).

Version 1.0.2 - 2000-10-16 (GV)
-------------
+ (IppvImplementation.cc) add explicit instantiations to satisfy
linker on DEC/OSF (SOC-SPR-1765).

Version 1.0.1 - 2000-10-13 (GV)
-------------
+ add parameter oalcheck (true by default). The newly created
summary file is passed on to the OAL. This guarantees that the
summary file can be parsed by the OAL. Also, a number of rough
sanity checks are performed: existence of attitude, time
correlation, and orbit file.
+ change name to the LOWER_THRESHOLD_Qy_CCDx parameters, as
requested by MPE. x=0,1,2 now.

Version 1.0 - 2000-10-06 (GV)
-----------
+ approved.
+ exposure start/end for OM now relies on SID=148/149 instead of
dates in DATE-OBS DATE-END.

Version 0.18 - 2000-09-29 (GV)
------------
+ (src/IppvTranslator.cc) in line with new Xmm::* names for the
PN-related filters.

Version 0.17 - 2000-09-29 (GV)
------------
+ (src/OdfObject.cc) also ask the cal server to be silent about
being unable to match the FILTER. This is required because the OM
data files have since ODS 10.2.3 a FILTER attribute, but this
attribute is the hardware value (a number) and not a cal-friendly
string.

Version 0.16 - 2000-09-26 (GV)
------------
+ (src/odfingest.cc) now remove ODF (SUM.ASC) file from the file
list to remove an ambiguity for the OAL.
+ (src/OdfObject.*) add removeFile(string) to implement the
above. Method icdosf() was private is now public.

Version 0.15 - 2000-09-25 (GV)
------------
+ review the instrument mode determination code.
+ doc updated and extended.

Version 0.14 - 2000-09-22 (GV)
------------
+ make instrument determination optional
(findinstrumentmodes=no). When mode determination is switched off
odfingest does not need a CIF.
+ make ippv extraction optional (usehousekeeping=no). When ippv
extraction is switched off odfingest does not need any HK files.
+ make use of new cal interface to set state without issuing a
warning, then intercept the condition mode->isUncertain(). Issue a
warning and set the mode to unknown.

Version 0.13 - 2000-09-19 (GV)
------------
+ implement PN filter translation from potentiometer readout to
string.
+ fix translation of PN fileter wheel name, as ODF ICD contained a
typo (FILTRA-THIN should have been FILTRA-THIN1). Expect more of
these.
+ when translation cannot be made, issue warning instead of
aborting.
+ Ippv class can now be told _not_ to check that a parameter value
remains constant throughout the exposure. This is used for the
EPIC clock RESET times: here we're interested in the value when
the exposure start and we don't care if the value changes during
the exposure, as this change is normal for exposures lasting more
than 2^15s.
+ Ippv: real32 version calculates the mean of the parameter over
the exposure instead of taking the first value. This is used to
calculate the mean position of the PN filter wheel. This value is
then translated to a string.

Version 0.12 - 2000-09-18 (GV)
------------
+ (src/IppvExtractorImplementation.cc) start looking for IPPVs
from tstart + 60s to tstop - 60s. This is to make sure that when
we look the relevant telemetry packets have been received.

Version 0.11 - 2000-09-18 (GV)
------------
+ (src/OdfObject.cc) improve camera mode, ccd mode determination
algorithm.

Version 0.10 - 2000-09-18 (GV)
------------
+ add CCD/camera mode string to the IPPV list
(via CAL, hence it now requires SAS_CCF to be set). Look for MODE
= in the summary file.
+ (config/odfingest.par) usecanonicalname=yes by default.

Version 0.9 - 2000-09-15 (GV)
-----------
+ add IppvTranslators, so that the MOS/OM filter (hw) positions
are translated to CAL-friendly strings. PN filter (string) is also
edited for typos.

Version 0.8 - 2000-09-14 (GV)
-----------
+ (src/IppvImplementation.cc) string Ippv were being extracted
incorrectly because the DAL does not report the correct number of
elements when a column segment is extracted. See comments in th
source for the details (look for FIXME).

Version 0.7 - 2000-09-07 (GV)
-----------
+ (doc/odfingest_description.tex) complete re-write with parameter
names, errors and warnings, housekeeping parameter names, output
file structure.

Version 0.6 - 2000-09-06 (GV)
-----------
+ add IPPV handling: all instruments are covered.

Version 0.5 - 2000-07-08 (GV)
-----------
+ disable test harness, as at present there are no compliant
ODFs to run on.

Version 0.4 - 2000-06-22 (GV)
-----------
+ sort file list (through copy, as it fails directly)
+ Remove extra line when copying PROPOSAL record.
+ Have both exposure-level and observation-level identifiers for
the exposures.

Version 0.3 - 2000-06-21 (GV)
-----------
+ absolute exposure identifier redefined to be [SU]nnn.

Version 0.2 - 2000-06-21 (GV)
-----------
+ first functional version.
+ implement IPPV support, however not activated because of shortcoming in the ODF HK datasets.

Version 0.1 - 2000-06-08 (GV)
-----------

+ Created. First version.