cansas1d binding IgorPro

From canSAS
Revision as of 15:46, 19 May 2008 by Jemian (talk | contribs) (show what metadata looks like)

canSAS 1-D SAS XML format binding to IgorPro

An import tool (binding) for IgorPro has been created (cansasXML.ipf). You can check out the IgorPro working directory from the SVN server (see instructions below).

To use the canSASxml.ipf procedure, you must have the XMLutils XOP IGOR plugin installed. See the Usage Notes below.

Note that the code described here is not a complete user interface. (See further comments below.) It is expected that this code will be called by a graphical user interface routine and that routine will handle the work of copying the loaded SAS data in IgorPro from the root:Packages:CS_XMLreader data folder to the destination of choice (including any renaming of waves as desired).

file cansasXML.ipf
author Pete R. Jemian <>
date 2008-05-19
version 1.07 (requires latest XMLutils XOP -- see below)
purpose implement an IgorPro file reader to read the canSAS 1-D reduced SAS data in XML files
adheres to the cansas1d/1.0 standard:
requires IgorPro:
XMLutils - XOP: (IGOR.5.04.x-1.x-dev, 2008-May-19)

Checkout of support code in Subversion

Subversion ( is a program for managing software versions. There are command line and GUI clients for a variety of operating systems. We won't recommend any here but will show the command lines necessary.

XMLutils XOP

The XMLutils XOP, written by Andrew Nelson (ANSTO), is hosted on the IgorExchange (

One good location to place the checked out XMLutils directory is in the Wavemetrics directory, next to the Igor Pro Folder

svn:// XMLutils

In the future, to retrieve an updated version of this support, go into the XMLutils directory (created above) and type the command

svn update

This will check the repository and update files as needed. If the installer program was updated, you;ll need to run the new installer program. It is not necessary to uninstall first.

The installer executables contained in the download will do all the installation for you. They will place the XOP in the folder /User Procedures/motofit/XMLutils, and create a shortcut/alias to the plugin in /Igor Extensions. Packages from other facilities should place the XOP there as well.


Check out the canSAS 1d SAS XML reader from the subversion repository:

svn checkout cansas-1dwg

This will download lots of extra files. The file of interest is in the IgorPro directory and is called cansasXML.ipf

In the future, to retrieve an updated version of this support, go into the cansas-1dwg directory (created above) and type the command

svn update

This will check the repository and update files as needed.


  1. License and Install IgorPro (should have already been done by now)
  2. Quit IgorPro if it is running
  3. Download XMLutils XOP. Either checkout from subversion (see above) or, with a web browser, go to
  4. Install XMLutils XOP by double-clicking the installer for you operating system.
  5. Download cansasXML.ipf. Either checkout from subversion (see above) or, with a web browser, copy cansasXML.ipf from on-line subversion repository:
  6. Copy cansasXML.ipf file to ...\Wavemetrics\Igor Pro Folder\User Procedures (or file system equivalent)
  7. Then, you should be able to restart IgorPro and progress from there

Usage Notes

To use the canSASxml.ipf procedure, you must have the XMLutils XOP IGOR plugin installed. This may be downloaded from the IgorExchange Project site. There are installer executables contained in the download that will do all the installation for you. Each installer will place the XOP in the folder ...\Wavemetrics\Igor Pro Folder\User Procedures\motofit\XMLutils, and create a shortcut/alias to the plugin in ...\Wavemetrics\Igor Pro Folder\Igor Extensions.

What it does

Given an XML file, CS_XmlReader(fileName) attempts to open the file and read its contents as if it conformed to the canSAS XML standard for reduced 1-D SAS data (cansas1d/1.0, also known as SASXML). If the file is found to be non-conforming, then CS_XmlReader(fileName) returns with an error code (show below), otherwise it returns 0 that indicates no error. All data read by this code is left in the IgorPro data folder root:Packages:CS_XMLreader for pickup by the calling routine. (Two examples are provided to show how a routine might retrieve the data.)

After opening the XML file (with a file identifier fileID), control is passed to CS_1i_parseXml(fileID) which then walks through the XML elements. For each SASentry in the file, a new data folder is created with the name derived from the Title element (or best effort determination). Efforts are taken to avoid duplication of data folder names (using standard IgorPro routines). For SASentry elements that contain more than one SASdata element, a SASdata folder is created for each and the corresponding I(Q) is placed in that subfolder. When only one SASdata is found, the I(Q) data is placed in the main Title folder.

data columns

Each column of data in the SASdata/Idata/* table is placed into a single IgorPro wave. At present, the code does not check for non-standard data columns. (The capability is built into the code but is deactivated at present).


Additional metadata is collected into a single text wave (metadata) where the first column is an identifier (or key) and the second identifier is the value. Only those keys with non-empty values are retained in the metadata table. CAUTION: The values are not checked for characters that may cause trouble when placed in a wave note. This will be the responsibility of the calling routine to clean these up if the need arises.

The code checks for most metadata elements and will check for repeated elements where the standard permits.

Here is an example of the metadata for the cs_collagen_full.xml data:

i (row) metadata[i][0] (key) metadata[i][1] (value)
0 xmlFile cs_collagen_full.xml
1 namespace cansas1d/1.0
2 Title dry chick collagen, d = 673 A, 6531 eV, X6B
3 Run Sep 19 1994 01:41:02 am
4 SASsample/ID dry chick collagen, d = 673 A, 6531 eV, X6B
5 SASinstrument/name X6B, NSLS, BNL
6 SASinstrument/SASsource/radiation X-ray synchrotron
7 SASinstrument/SASsource/wavelength 1.898
8 SASinstrument/SASsource/wavelength/@unit A
9 SASinstrument/SASdetector/@name X6B PSD
10 SASnote
Sep 19 1994     01:41:02 am     Elt: 00090 Seconds 
ID: No spectrum identifier defined
Memory Size: 8192 Chls  Conversion Gain: 1024  Adc Offset: 0000 Chls

dry chick collagen, d = 673 A
6531 eV, X6B

XML foreign namespace elements

These are ignored at this time.

XML namespace and header

The routine does a best-efforts check to ensure that the given XML file conforms to the required XML file header. If you take a minimalist view (a.k.a. a shortcut), it is likely that your file may be refused by this and other readers. Pay particular attention to UPPER/lower case in the text cansas1d/1.0 as this is a key component used to index through the XML file.

List of Functions

These are (most of) the FUNCTIONS in the cansasXML.ipf code. The only functions of interest are CS_XmlReader(fileName) which reads the named XML file and and loads SAS data and the two demonstration functions prj_grabMyXmlData() and prjTest_cansas1d() that together show a usage example.

open a canSAS 1-D reduced SAS XML data file
  • input: fileName (string) name of canSAS XML file (can include file system path name to file)
  • returns:
    • 0 successful
    • -1: XML file not found
    • -2: root element is not <SASroot> with valid canSAS namespace
    • -3: <SASroot> version is not 1.0
    • -4: no <SASentry> elements (NOT USED NOW)
    • -5: XOPutils needs upgrade
This is what guides the work, given a file ID returned from XMLOpenFile(), parses that file for SAS data and metadata
(1i in the function name signifies this is a function that supports INPUT from version 1.0 XML files)
CS_1i_getOneSASdata(fileID, Title, SASdataPath)
harvest the data and metadata in the specific SASdata element
harvest just one column (vector) of data
CS_1i_GetReducedSASdata(fileID, SASdataPath)
grab the data and put it in the working data folder
CS_1i_locateTitle(fileID, SASentryPath)
determine the title for this experiment
CS_appendMetaData(fileID, key, xpath, value)
queries XML file for xpath. If value is not empty, appends it to metadata where last is the new last row: metadata[last][0]=key; metadata[last][1]=value
CS_buildXpathStr(prefix, value)
this function can be used only with very simple XPath constructions
given a proposal string, returns a candidate folder name for immediate use
looks for element index in structure W_ElementList returned from call to XmlElemList(fileID)
returns the string containing the default namespace for the XML file
Builds a table of all namespaces used in the XML file and appends W_ElementList with full namespace-xpath string for each element.
CS_simpleXmlListXpath(fileID, prefix, value)
Calls XMLlistXpath() with proper namespace prefix attached.
CS_simpleXmlWaveFmXpath(fileID, prefix, value)
Calls XMLwaveFmXpath() with proper namespace prefix attached.
CS_updateWaveNote(wavName, key, value)
adds (or replaces) definition of key=value in the wave note of wavName
CS_XmlStrFmXpath(fileID, prefix, value)
Calls XmlStrFmXpath() with proper namespace prefix attached. Trims the result string.
this function adds namespace info as necessary to simpleStr (an XPath)
Calls TrimWSL(TrimWSR(str))
Trims white space from left (leading) end of str
Trims white space from right (trailing) end of str
Demonstration function that calls CS_XmlReader(fileName) for many of the test data sets.
Demonstration function that moves loaded data from root:Packages:CS_XMLreader to a user's data folder. (In this example, that folder is root:PRJ_canSAS.)
Demonstration function that reads an ISIS/LOQ file and copies the data to the root folder a la COLLETE