DAP4: Constraint and Query - Revision history

Ndp: /* Constraint Expression */

2014-02-08T21:03:13Z

Constraint Expression

← Older revision		Revision as of 21:03, 8 February 2014
Line 33:		Line 33:
	The DAP4 CE should be held in a single key value pair. The CE value should be URI encoded** as appropriate for it's character content and all occurrences of the "'''&'''" must also be URI encoded. Combined with a riff on [[DAP4:_DAP4_Filter_Constraints \| Dennis' Filter Proposal ]] we could get to something like this:		The DAP4 CE should be held in a single key value pair. The CE value should be URI encoded** as appropriate for it's character content and all occurrences of the "'''&'''" must also be URI encoded. Combined with a riff on [[DAP4:_DAP4_Filter_Constraints \| Dennis' Filter Proposal ]] we could get to something like this:

	: <~~font size="3"><code~~> ?dap4.ce=n,a[3:1:100],k\|x<=y\|z=y</~~code></font~~>		: <tt> ?dap4.ce=n;a[3:1:100];k\|x<=y\|z=y</tt>

	Where we are asking for the variable '''n''', the 3rd through the 100th elements of the array '''a''', and all members of the rows of sequence '''k''' for which '''k.x<=k.y''' and '''k.z=k.y'''		Where we are asking for the variable '''n''', the 3rd through the 100th elements of the array '''a''', and all members of the rows of sequence '''k''' for which '''k.x<=k.y''' and '''k.z=k.y'''

Ndp: /* Examples */

2014-02-08T21:01:37Z

Examples

← Older revision		Revision as of 21:01, 8 February 2014
Line 68:		Line 68:

	Example:		Example:
	: <~~font size="3"><code~~>?CE=constraintExpressionString&d4.checksums=off&d4.arraySelection=mask(0x00)&d4.saveAs=newgranule&d4.async=true</~~code></font~~>		: <tt>?dap4.ce=constraintExpressionString&dap4.checksums=off&dap4.arraySelection=mask(0x00)&dap4.saveAs=newgranule&dap4.async=true</tt>


	Example:		Example:
	: <~~font size="3"><code~~>?CE=build_data(place,u[100:200])&d4.checksums=off&d4.arraySelection=Sequence&d4.saveAs=newgranule&d4.asyncAccepted</~~code></font~~>		: <tt>?dap4.ce=build_data(place,u[100:200])&dap4.checksums=off&dap4.arraySelection=Sequence&dap4.saveAs=newgranule&dap4.asyncAccepted</tt>

	== Rationale for the solution ==		== Rationale for the solution ==

Ndp: /* Extensibility */

2014-02-08T20:59:50Z

Extensibility

← Older revision		Revision as of 20:59, 8 February 2014
Line 55:		Line 55:
	=== Extensibility ===		=== Extensibility ===

	We ~~should~~ claim a set of "~~keys~~" to be owned by the DAP4 protocol. ~~I think they should be:~~		We claim the set of all query string keys beginning with the string "<tt>dap4.</tt>" to be owned by the DAP4 protocol.

	* "~~CE" and~~/~~or "ce~~" for the constraint expression.		Examples:
	* "async" for asynchronous response control.
	* ~~And anything~~ beginning with "d4." to give us future expandability~~. We might want to explicitly define include "d4.ce" and "d4.async" as synonyms of "CE" and "async" in the first draft~~.		* "<tt>dap4.ce</tt>" for the constraint expression.
			* "<tt>dap4.async</tt>" for asynchronous response control.
			* Any other key beginning with "<tt>dap4.</tt>" to give us future expandability.

	This will allow us to add service controls and inputs over time without the risk of interfering with other efforts to add service controls and inputs to various DAP4 servers.		This will allow us to add service controls and inputs over time without the risk of interfering with other efforts to add service controls and inputs to various DAP4 servers.

Ndp: /* Additional Server Controls */

2014-02-08T20:55:52Z

Additional Server Controls

← Older revision		Revision as of 20:55, 8 February 2014
Line 48:		Line 48:
	=== Additional Server Controls ===		=== Additional Server Controls ===

	By removing the "'''&'''" character from the CE we can now use the rest of the ~~QS to be used~~ to pass server controls and options. Consider if we were to take my corollary to Dennis' filter proposal and allow filters to be applied to arrays. The resulting data object could be expressed as a Sequence, or as a array with a mask applied. A server could support both and the type of result might be controlled by a KVP. You might even specify the mask value:		By removing the "'''&'''" character from the CE we can now use the rest of the query string to pass server controls and options. Consider if we were to take my corollary to Dennis' filter proposal and allow filters to be applied to arrays. The resulting data object could be expressed as a Sequence, or as a array with a mask applied. A server could support both and the type of result might be controlled by a KVP. You might even specify the mask value:
	: <font size="3"><code>?CE=u[25:1:1000][75:1:1000]\|u>4.0\|u<12.7&d4.arraySelection=mask(0x00)</code></font>		: <font size="3"><code>?dap4.ce=u[25:1:1000][75:1:1000]\|u>4.0\|u<12.7&dap4.arraySelection=mask(0x00)</code></font>
	Or, ask for a Sequence:		Or, ask for a Sequence:
	: <font size="3"><code>?CE=u[25:1:1000][75:1:1000]\|u>4.0\|u<12.7&d4.arraySelection=Sequence</code></font>		: <font size="3"><code>?dap4.ce=u[25:1:1000][75:1:1000]\|u>4.0\|u<12.7&dap4.arraySelection=Sequence</code></font>

	=== Extensibility ===		=== Extensibility ===

Ndp: /* Constraint Expression */

2014-02-08T20:54:38Z

Constraint Expression

← Older revision		Revision as of 20:54, 8 February 2014
Line 33:		Line 33:
	The DAP4 CE should be held in a single key value pair. The CE value should be URI encoded** as appropriate for it's character content and all occurrences of the "'''&'''" must also be URI encoded. Combined with a riff on [[DAP4:_DAP4_Filter_Constraints \| Dennis' Filter Proposal ]] we could get to something like this:		The DAP4 CE should be held in a single key value pair. The CE value should be URI encoded** as appropriate for it's character content and all occurrences of the "'''&'''" must also be URI encoded. Combined with a riff on [[DAP4:_DAP4_Filter_Constraints \| Dennis' Filter Proposal ]] we could get to something like this:

	: <font size="3"><code> ?dap4=n,a[3:1:100],k\|x<=y\|z=y</code></font>		: <font size="3"><code> ?dap4.ce=n,a[3:1:100],k\|x<=y\|z=y</code></font>

	Where we are asking for the variable '''n''', the 3rd through the 100th elements of the array '''a''', and all members of the rows of sequence '''k''' for which '''k.x<=k.y''' and '''k.z=k.y'''		Where we are asking for the variable '''n''', the 3rd through the 100th elements of the array '''a''', and all members of the rows of sequence '''k''' for which '''k.x<=k.y''' and '''k.z=k.y'''

Ndp: /* Examples */

2012-11-13T20:09:18Z

Examples

← Older revision		Revision as of 20:09, 13 November 2012
Line 66:		Line 66:

	Example:		Example:
	: <font size="3"><code>?CE=constraintExpressionString&d4-checksums=off&d4-arraySelection=mask(0x00)&d4-saveAs=newgranule&d4-async=true</code></font>		: <font size="3"><code>?CE=constraintExpressionString&d4.checksums=off&d4.arraySelection=mask(0x00)&d4.saveAs=newgranule&d4.async=true</code></font>


	Example:		Example:
	: <font size="3"><code>?CE=build_data(place,u[100:200])&d4-checksums=off&d4-arraySelection=Sequence&d4-saveAs=newgranule&d4~~-async=true~~</code></font>		: <font size="3"><code>?CE=build_data(place,u[100:200])&d4.checksums=off&d4.arraySelection=Sequence&d4.saveAs=newgranule&d4.asyncAccepted</code></font>

	== Rationale for the solution ==		== Rationale for the solution ==

Ndp: /* Additional Server Controls */

2012-11-13T20:08:21Z

Additional Server Controls

← Older revision		Revision as of 20:08, 13 November 2012
Line 49:		Line 49:

	By removing the "'''&'''" character from the CE we can now use the rest of the QS to be used to pass server controls and options. Consider if we were to take my corollary to Dennis' filter proposal and allow filters to be applied to arrays. The resulting data object could be expressed as a Sequence, or as a array with a mask applied. A server could support both and the type of result might be controlled by a KVP. You might even specify the mask value:		By removing the "'''&'''" character from the CE we can now use the rest of the QS to be used to pass server controls and options. Consider if we were to take my corollary to Dennis' filter proposal and allow filters to be applied to arrays. The resulting data object could be expressed as a Sequence, or as a array with a mask applied. A server could support both and the type of result might be controlled by a KVP. You might even specify the mask value:
	: <font size="3"><code>?~~dap4~~=u[25:1:1000][75:1:1000]\|u>4.0\|u<12.7&arraySelection=mask(0x00)</code></font>		: <font size="3"><code>?CE=u[25:1:1000][75:1:1000]\|u>4.0\|u<12.7&d4.arraySelection=mask(0x00)</code></font>
	Or, ask for a Sequence:		Or, ask for a Sequence:
	: <font size="3"><code>?~~dap4~~=u[25:1:1000][75:1:1000]\|u>4.0\|u<12.7&arraySelection=Sequence</code></font>		: <font size="3"><code>?CE=u[25:1:1000][75:1:1000]\|u>4.0\|u<12.7&d4.arraySelection=Sequence</code></font>

	=== Extensibility ===		=== Extensibility ===

Ndp: /* Examples */

2012-11-12T23:48:19Z

Examples

← Older revision		Revision as of 23:48, 12 November 2012
Line 66:		Line 66:

	Example:		Example:
	: <font size="3"><code>?~~dap4~~=CE&checksums=off&arraySelection=mask(0x00)&saveAs=newgranule&async=true</code></font>		: <font size="3"><code>?CE=constraintExpressionString&d4-checksums=off&d4-arraySelection=mask(0x00)&d4-saveAs=newgranule&d4-async=true</code></font>


	Example:		Example:
	: <font size="3"><code>?~~dap4~~=build_data(place,u[100:200])&checksums=off&arraySelection=Sequence&saveAs=newgranule&async=true</code></font>		: <font size="3"><code>?CE=build_data(place,u[100:200])&d4-checksums=off&d4-arraySelection=Sequence&d4-saveAs=newgranule&d4-async=true</code></font>

	== Rationale for the solution ==		== Rationale for the solution ==

Ndp: /* Constraint Expression */

2012-11-12T23:39:36Z

Constraint Expression

← Older revision		Revision as of 23:39, 12 November 2012
Line 31:		Line 31:
	=== Constraint Expression ===		=== Constraint Expression ===

	The DAP4 CE should be held in a single key value pair. ~~This means that~~ the "'''&'''" ~~character~~ must ~~not~~ be ~~used in the content of the CE~~. Combined with a riff on [[DAP4:_DAP4_Filter_Constraints \| Dennis' Filter Proposal ]] we could get to something like this:		The DAP4 CE should be held in a single key value pair. The CE value should be URI encoded** as appropriate for it's character content and all occurrences of the "'''&'''" must also be URI encoded. Combined with a riff on [[DAP4:_DAP4_Filter_Constraints \| Dennis' Filter Proposal ]] we could get to something like this:

	: <font size="3"><code> ?dap4=n,a[3:1:100],k\|x<=y\|z=y</code></font>		: <font size="3"><code> ?dap4=n,a[3:1:100],k\|x<=y\|z=y</code></font>

	Where we are asking for the variable '''n''', the 3rd through the 100th elements of the array '''a''', and all members of the rows of sequence '''k''' for which '''k.x<=k.y''' and '''k.z=k.y'''		Where we are asking for the variable '''n''', the 3rd through the 100th elements of the array '''a''', and all members of the rows of sequence '''k''' for which '''k.x<=k.y''' and '''k.z=k.y'''




			<nowiki>**</nowiki> In general URIs as defined by [http://tools.ietf.org/html/rfc3986 RFC 3986] may contain any of the following characters:

			<pre>ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-._~:/?#[]@!$&'()*+,;=.</pre>

			Any other character needs to be encoded with the percent-encoding (%hh). Each part of the URI has further restrictions about what characters need to be represented by an percent-encoded word.

	=== Additional Server Controls ===		=== Additional Server Controls ===

Ndp: /* Discussion */

2012-11-07T00:44:50Z

Discussion

← Older revision		Revision as of 00:44, 7 November 2012
Line 159:		Line 159:
	# The browser escapes the URL, curl does not (although you may have to shell escape components of a URL/CE to get command-line-curl and your shell to swallow it)		# The browser escapes the URL, curl does not (although you may have to shell escape components of a URL/CE to get command-line-curl and your shell to swallow it)
	# If you utilize URL encoding to encode a '&' sign in the query string, then Tomcat will ignore it when breaking the query string into KVPs, but will subsequently decode it before generating the keys and their values.		# If you utilize URL encoding to encode a '&' sign in the query string, then Tomcat will ignore it when breaking the query string into KVPs, but will subsequently decode it before generating the keys and their values.
	# Every '&' in the query string received by Tomcat that is not encoded (i.e. that appears as itself in the query string) will signal the Tomcat URL parser to start to generate a new KVP. I say start because there could be two adjacent ampersands (&&), in which case they are treated as a single ampersand. Additionally, the content may lack an equals symbol or a value, in which case the entire substring (between ampersands or between the last ampersand and the end of the query string) becomes a key whose value is the empty string. <br/> This I think means that we might also consider using simple keywords to control server behavior if we wished - for example instead of saying: <font size="3"><code>&async=true </code></font> we might instead say simply <font size="3"><code> &acceptAsync </code></font> with the server simply testing for the presence (or lack thereof) of the particular key.<br/>		# Every '&' in the query string received by Tomcat that is not encoded (i.e. that appears as itself in the query string) will signal the Tomcat URL parser to start to generate a new KVP. I say start because there could be two adjacent ampersands (&&), in which case they are treated as a single ampersand (and Tomcat generates a warning: "WARNING: Parameters: Invalid chunk '' ignored."). Additionally, the content may lack an equals symbol or a value, in which case the entire substring (between ampersands or between the last ampersand and the end of the query string) becomes a key whose value is the empty string. <br/> This I think means that we might also consider using simple keywords to control server behavior if we wished - for example instead of saying: <font size="3"><code>&async=true </code></font> we might instead say simply <font size="3"><code> &acceptAsync </code></font> with the server simply testing for the presence (or lack thereof) of the particular key.<br/>
	All of which pretty much confirms that Dennis' point regarding Parsing and Escaping of the DAP4 URL is spot on, and that Tomcat does exactly what he suggests. I believe that the specification should make clear that clients need to URL escape ( %hh where hh is the hexadecimal value of the character being escaped)(uh... UNICODE?) the keys and values but not the ampersands that seperate them when making requests to DAP4 servers. <br/>		All of which pretty much confirms that Dennis' point regarding Parsing and Escaping of the DAP4 URL is spot on, and that Tomcat does exactly what he suggests. I believe that the specification should make clear that clients need to URL escape ( %hh where hh is the hexadecimal value of the character being escaped)(uh... UNICODE?) the keys and values but not the ampersands that seperate them when making requests to DAP4 servers. <br/>
	How does that sound?<br/>		How does that sound?<br/>