GSPARC: User guide section

Sample and Data Naming Conventions

Each item (sample) in the physical archive is assigned a unique identification code. This code must comply with the GSPARC naming conventions described below. Some flexibility in expressing the identification code is allowed and is somewhat dependent on context (see later).The fundamental identification string consists of three essential fields plus optional extensions. The generic expression (shown here in standard software notation) is as follows :-

<family ID><d>[<format ID><d>]<itemID>[<d><target ID><d><target type ID>[<d><expt date ID><expt seq ID><inst ID>]]

where < > is a field, d is a field delimiter and [ ] is an optional extension

FAMILY ID: Type NUMERIC. This number must be unique to an associated group (or family) of samples. Usually a family of samples is simply a set of items derived from specimens obtained from a common source (i.e. a particular rock, outcrop, horizon etc). However, if there is sufficient variation in properties to jeopardise "commonality" then a separate family must be created. In such cases relationships and associations with other "families" can still be maintained via other assigned attributes. Note that is possible for a family to consist of a single item thus making this identifier unique in the database. A single parameter search of the database using this identifier would return all items (samples) contained within this family. Note that it is possible to pad this field with leading zeros if this is appropriate (i.e. filenames may benefit from this).

FORMAT ID:Type CHARACTER*2. A pair of characters selected from a lookup table describing format of the specimen. See reference section for full list of options. In some circumstances this field can be omitted (see example below). A two parameter search of the database based on the format ID plus the family ID would return all items (samples) in a family of this format type.

ITEM ID:Type NUMERIC. A number unique to the database (i.e. can occur once only). Searching the GSPARC database for this parameter would always return this particular item (sample) irrespective of other attributes. Note that it is possible to pad this field with leading zeros as in the case of the family ID.

DELIMITER: Type CHARACTER*1. Preferences are period (.) or dash (-) but can be any (sensible) non-alphanumeric symbol. If this delimiter character is omitted then the format identifier MUST be included to delimit the two essential numeric fields (for obvious reasons). A mix of delimiters in the same identification code is allowed (although not recommended).

TARGET ID: Type ALPHANUMERIC*N where N is less than 4. Any combination of characters and/or numerics to identify a feature or target on the sample. Note that each target must have a unique ID to avoid ambiguity.

TARGET TYPE ID:Type CHARACTER*2. A pair of characters selected from a lookup table describing the nature of the target. Examples include sawn face (SF), polished face (PF), weathered exterior (WE), removed fragment (XX) and so on. A full list is given in the reference section.

EXPT DATE ID: Type NUMERIC*6. Date of which the experiment was performed, initiated or completed. Strictly date format yymmdd.

EXPT SEQ ID: Type ALPHANUMERIC*N where N is usually 1 but can be >1 (recommend keep as short as possible). An incremental character field pertaining to the experiment performed that day (i.e. A would be first, B second and so on up to Z, thereafter A1, B1 and so on).

INST ID: Type ALPHANUMERIC*N where N is variable (recommend keep as short as possible). Any combination of characters and/or numerics to identify instrument used to perform experiment. A full list is given in the reference section.

Example:

Consider a sample with a family ID 29, a format type hand specimen (i.e. HS) and an item ID 49. It is entirely feasible to express this item as 29.49 if it obvious that the sample is a hand specimen (i.e. in the case of an image caption) or as 29.HS.49 (or 29HS49) if the format is relevant or not obvious (i.e. when describing data for siblings of different formats).

The sample also has four distinctive targets of interest, a fresh naturally cleaved surface, a sawn face, a weathered face and a fragment that was removed for microscopic analysis. Each of these could be expressed as 29.49.01.FS, 29HS49-02-SF, 0029.HS.0049.B.WE and 29.49-4-XX respectively. Note that in the case of the fragment, the area it originated from could be expressed as 29.49-4-FS in the case of a residual fresh surface or 29.49-4-WI if an internal weathered surface is left.

Say on the 1st of February 2004 a set of images were acquired of these samples using the Beagle2 qualification model microscope. These were the 4th and 5th image sets acquired on that day with this instrument, the first three being of another sample being studied. Two sub-directories would be created in the instrument directory B2QMMIC called 040201D-29HS49-01-FS and 040201E-29HS49-B-WE. To complement the microscope images (which incidently use an instrument specific naming convention, something entirely valid within GSPARC), witness images of each experimental setup were acquired using a standard digital camera. Each witness image could be called setup.jpg as long as both reside in the relevant sub-directory defined above. However, if these images where to be used elsewhere (i.e. in the shared image archive for example) there full (unambiguous) names should be used (i.e. 040201D-29HS49-01-FS-setup.jpg and 040201D-29HS49-B-WE-setup.jpg).

Note the above example is for the most part ficticious and is for illustrative purposes only.