Scheme, count

Agreements

Line

Fasten broken line to objects

Let's look at typical situation of drafting database scheme.

We put "City" and "Street" in cells of invisible table to set them in page.

It's required to fasten a broken line to tables.

I offer to enter tags line , ld (origin of designations: ld = line division) and pack with attribute host , which equal to attribute id of tag td.

<table>
  <tr>
    <td></td>
    <td>
      <table id="id1">
        <caption> Street      </caption>
        <tr><td>  idStreet    </td></tr>
        <tr><td>  idCity (FK) </td></tr>
        <tr><td>  name        </td></tr>
      </table>
    </td>
  </tr>
  <tr>
    <td>
      <table id="id2">
        <caption> City   </caption>
        <tr><td>  idCity </td></tr>
        <tr><td>  name   </td></tr>
      </table>
    </td>
    <td></td>
  </tr>
</table>

<line>
  <pack host="id1" align="side">
  <pack host="id2" dot="last" align="side">
  <ld x="0" y="0"></ld>
  <ld x="0" y="50"></ld>
  <ld x="100" y="50"></ld>
  <ld x="100" y="100"></ld>
</line>

Broken line is scaled during fastening to 2 objects.

Tag pack ties a point of broken line to object (in this case - to table). Identifier id (if it's absent - serial number, counted from one) of a tied point is specified in attribute dot. Special values of attribute dot are first and last, which specifies, that first or last point of broken line should be tied.

It's possible to fasten broken line (using tag pack) not only for the ends, but also for two intermediate points, or for end and intermediate points.

If there is no ld inside tag line, then objects are connected by one direct line.

Alignment

Characteristic align of tag pack specifies, where the end of tied section will be located, and accepts one of the following values: side (default value), center, into. Dotted line shows an invisible part of a section.

side	center	into

Half-tied broken line

If broken line has only one tag pack, then broken line is not scaled and is tied to object only by its one point. In this case broken line is drawn so, that its axis "Ox" goes horizontal from left to right, and its axis "Oy" goes vertically from below to top.

coordinates line, consist of one section	location of line in html-document

Not tied broken line

<pic>
  <line> ... </line>
</pic>

pic {
  reper=100.5 200.5;
  range=200.5 400.5;
}

Put object into point of broken line

Any objects can be located in a broken line point (we name their enclosed objects further)

• text: <ld> it's a text </ld>
• any tag structure:
• table <ld> <table> ... </table> </ld>
• figure <ld> <img src= > </ld>
• etc.

Location of object centre comparatively line point is still specified in characteristic align, but this characteristic must be used in tag ld. Putting object into point of broken line allows to make complex constructions

and putting in cascade - to make very complex constructions, for example database schemes

Case, when object, located in a point, may cover (including cover after drag-and-drop) other objects, is regulated by characteristic push of tag ld. Value yes in any tag ld means, that objects, intersecting with it, must be moved automatically by browser in a direction, which is clear in the following pictures. Default value is no.

Note for browser developers.

It's bad to allow any objects to have attributes "x", "y", and reason is following. All objects should be grouped into layers (i.e. into broken lines), and even such objects, which are a dump, should be placed inside invisible broken line (you can name this broken line as "heap", "dump", etc.).

Explanatory inscriptions

Explanatory inscriptions for points and sections are specified in attributes point-note and section-note of tag ld. Location of inscription is clear in the following pictures.

If you need to put any text inside broken line (more precisely: in geometrical centre of points of broken line), you must put tag contour with attribute text inside tag line. Tag contour has characteristics color, white-space and all characteristics of group font.

contour { color=lime }	<line> <ld> ... <ld> <contour text="open"> </line>
contour { white-space=pre }	<line> <ld> ... <ld> <contour text="p1 7.28 p2 1.93 p3 5.02"> </line>

Closed, painted over

Assignment any colours to characteristic background-color of tag contour results to painting over a line by the specified colour (it's convenient to write minerals, chisels, military objects on a map). Default value null means, that value, specified in surrounding tag, is applied.

Only internal area of self-crossing line are painted over.

If two polygons (closed, painted over broken lines) are intersected in result of server or user command, than a place of their intersecting are painted over by "average" colour.

Count can have several tags contour. Assignment a colour to characteristic background-color paint over only one contour of count.

Roalover

Help, which pop-up, when mouse is over points and section, is defined in attributes point-title and section-title of tag ld. If roalover help for a point is not specified, then browser shows coordinates of point as roalover help.

Fasten broken line to broken line

It's necessary to specify identifier id (if it's absent - serial number) of point of other broken line in attribute ptr of tag pack for that. Special values of attribute ptr are first and last, which specifies, that it's need to fasten to first or last point of other broken line.

If broken line has only one tag pack, i.e. is half-tied, then it can be rotated by specification a new location in attribute angle of tag pack.

	angle=median angle=m
	angle=antimedian angle=am

Resume

Database support

Geographical maps, GML

Tag <link type="point"> can be inside of tag line. If file "a.txt" has following contents

0 0
0 50
100 50
100 100

the example in article beginning can be written so

<line>
 <link src="./a.txt" type="point">
 <pack host="id1" align="side">
 <pack host="id2" dot="last" align="side">
</line>

You find advantage for figure, like given below.

<line>
 <link src="./a.txt" type="point">
</line>

In third field there is - specially for a forms with broken line - unique identifier of broken line point (usual, not necessarily it is number): primary key or serial number.

0 0 7
0 50 h21
100 50 5
100 100 8

Fourth field prolongs to end of string and contains (probably through a blanks) html-tags (or text), which should be placed in a point.

Counts

<line>
 <link src="./a.txt" type="point">
 <link src="./b.txt" type="section">
 <pack host="id1" align="side">
 <pack host="id2" dot="last" align="side">
</line>

72  250 4       NH
317 182 anyword OH
581 42  7       Na <sup> + </sup>

4 7 1  triple red
4 5 ab wedge

Second table specifies count bonds and contains records of following fields: unique identifier of initial point, unique identifier of final point, unique identifier of section, fourth field prolongs to end of string and contains values of several section characteristics through a blanks (browser understand name of section characteristic by value of characteristic).

Presence of tag <link type="section"> means, that bonds will be drawn not by way of following.

Standard examples of count are a structural chemical formulas.

Explanatory inscriptions

Explanatory inscriptions of beginning and end of sections are defined in characteristics section-parent and section-child of tag ld.

ld {
  section-parent: "parent";
  section-child:  "child";
}

Explanatory inscriptions inside contours

If it's necessary to put explanatory inscriptions in any contours

<line>
 <contour src="./c1.txt" text="open">
 <contour src="./c2.txt" text="any">
 <contour src="./c3.txt" text="some">
</line>

File of contour points is specified as one-column-table in a textual file, file name is specified in attribute src of tag contour. Points identifiers are listed consistently in this column. It's unimportant, whether any section passes between two next points or not - inscription are in geometrical centre of points.

2 
4 
7 
8 
11 
15 

If tag <link type="section"> is absent, then attribute src of first tag contour is ignored (contour is whole broken line), and rest tags contour are ignored.

Resume

Characteristics

Each tag ld (except for last) responds both for a point and for a section (last tag ld - only for a point).

All characteristics can be specified both for tag ld , and for tag line .

Characteristic color specifies text colour in a point and equal black by default.

Group of characteristics POINT

When contents of tag ld are empty, its characteristic point-type specifies, what geometrical figure will be drawn in a point (this charactetistic is similarly to characteristic list-style-type). Usual values of point-type are: none (default value), disc, circle, square, decimal. I offer to add: arrow (arrow in section end), barrow (backarrow, arrow in section beginning), botharrow (arrows in two section ends), star (asterisk), romb (rhombus), triangle (triangle), hex (hexagon), which will be necessary for writing scientific diagrams. And browser always draws whole (not half ! ) geometrical figure.

Characteristic point-color specifies colour of geometrical figure in a point. By default it's equal black.

Group of characteristics SECTION

Characteristic section-color specifies colour of broken line section. By default it's equal black.

Characteristic section-width specifies thickness of section in millimeters (fractional part of number is separated from natural by mark "point") and accepts also special values thin (default value), medium, thick.

Characteristic section-type specifies type of broken line section and accepts values: null (none) - invisible section, dotted, dashed, solid (default value), double (may be railway), groove, ridge. I offer to add value single as equal to solid, and also values wedge, bwedge, dwedge, bdwedge for drawing any out-going and in-coming line (they are necessary each time, when it's necessary to show, that some lines leave a plane of figure).

	triple
	wedge (if section-type=thin, then thin-to-medium, if section-type=medium, then medium-to-thick, etc)
	bwedge (backwedge, if section-type=thin, then medium-to-thin, if section-type=medium, then thick-to-medium, etc)
	dwedge (dottedwedge, wedge & dotted)
	bdwedge (backdottedwedge, backwedge & dotted)

Characteristics section are similar to characteristics of border group. And they are not depend of characteristic background-color.

Resume

Characteristics

tag	characteristic	default value	list of possible values
ld ( line )	point-type	null	null (none), disc, circle, square, decimal, arrow, barrow, botharrow, star, romb, triangle, hex
	point-color	black	black, maroon, green, olive, navy, purple, teal, gray, silver, red, lime, yellow, blue, fuchsia, aqva, white, gold, orange, violet, indigo, #FFFFFF
	color	- || -	- || -
	section-color	- || -	- || -
	section-type	solid	null (none), dotted, dashed, single (solid), double, groove, ridge, triple , wedge , bwedge , dwedge , bdwedge
	section-width	thin	thin, medium, thick, float

Forms with lines and maps

Tag line can be located inside tag form. Behaviour of broken lines under commands is the same as behaviour of all other html-objects, but only lines have:

• drag-and-drop
• 7 new parameters (attributes) in sending XML-text (action): num, x, y, dx, dy, user-x, user-y
• attributes point-menu and section-menu, which define menus for a point and section accordingly, are used together with attribute menu

Point and section of same unit can be written by two different styles after command, that is completely normal.

Styles specifications may be empty. It's necessary when equally represented broken lines are different semantically.

line.a {}
line.b {}

It's important to remember, that a point styles with the same name may be not equivalent.

line.a ld.x {color:red}
line.b ld.x {color:blue}

New points �oordinates are sent in parameters x and y. Summary (total) displacement of point are sent after many movings.

Last moved point (and object in it) or section lay over all others (inside tag layer, certainly).

Assignment value yes to characteristic user-xy of tag ld results to sending two additional attributes user-x and user-y in action section.

Drag-and-drop of object, enclosed in a point, is drag-and-drop of point.

If an enclosed object has changed style under commands, then tag ld does not change style.

• parameter name contains value of attribute name (and if it's absent - of attribute id) of tag line - further parameter value is designated by letter "A".
• parameter num contains serial number of ld (is counted from 1) or primary key of point or section - further parameter value is designated by letter "N".

Origin of designations: num = number, os = old style (state), ns = new style (state), cmd = command.

Moment of sending is described in separate document.

Work with points

click

Click on a point sends XML-text

<dot name=A cmd=click os=... ns=... num=N>

Usually programmer defines style ns in css-file for this case (form is sent on pressing button "submit"), because click on a point usually means a choice of a point.

There exist characteristic deformation for tag ld, by default equal yes. If it's no, then drag-and-drop only one point is impossible - two adjacent (neighbouring) sections go together, without changing angle between them.

<dot name=A cmd=double os=... ns=... num=N>

Click by right key ("right" click) on a point means occurrence menu for the point, in which user chooses a new point states (new style for a point). The XML-text is sent

<dot name=A cmd=... os=... ns=... num=N>

Point menu is set in atribute point-menu of tags ld or line.

If style ns is not found in css-file, then sense of menu element is another - that is choice of some operation for a point.

drag

<dragdot name=A num=N x=... y=...>

Drag-and-drog is very convenient to guess of points coordinates. If points coordinates are given in a separate file

<line>
  <link src="./a.txt" type="point">
</line>

If a points coordinates are in html-document, then new coordinates of drag-and-drop'ed points can be found out during sending them from a form.

If point has got upon other point (of this or other broken line, of this or other picture) - further we shall name it as accepting point, then browser sends additional XML-text

<dropdot name=A num=N onname= onnum= >

drop

Look at previous paragraph

<dragdot name=A1 num=N1                             x=  y=  >
<dot     name=A1 num=N1                    cmd=drag os= ns= >
<dropdot name=A1 num=N1 onname=A2 onnum=N2>
<dot                    name=A2   num=N2   cmd=drop os= ns= >

Work with sections

click

Click on a section sends following XML-text for the section point

<section name=A cmd=click os=... ns=... num=N>
<section name=A cmd=click os=... ns=... num=N user-x=... user-y=...>

Usually programmer defines style ns in css-file for this case (form is sent on pressing button "submit"), because click on a section usually means a choice of a section.

double

Double-click on a section sends XML-text for the section point

<section name=A cmd=double os=... ns=... num=N>
<section name=A cmd=double os=... ns=... num=N user-x=... user-y=...>

Usually programmer does not define style ns in css-file for this case (form is sent immediately), because double-click on a section usually means cutting it and creation new point in a place of click. Immediately occurs same, but updated html-page with a new point.

right-click

Click by right key ("right" click) on a section means occurrence menu for the section, in which user chooses a new section states (new style for a section). The XML-text is sent

<section name=A cmd=... os=... ns=... num=N>
<section name=A cmd=... os=... ns=... num=N user-x=... user-y=...>

Section menu is set in atribute section-menu of tags ld or line.

If style is not found in css-file, then sense of menu element is another - that is choice of some operation for a section.

drag

Section goes during drag-and-drop by parallel carrying. Form sends new coordinates of two points.

<dragdot name=A num=N   x=... y=...>
<dragdot name=A num=N+1 x=... y=...>

Two variants are possible, depending on value of characteristic move of tag ld :

• Moving of section end is shift point A into point B, move=dot (default value)
• Moving of section end is shift point A into point C, move=cross

dot	cross

If two points are combined into one, then second point (on following inside tag line) receives coordinates (null, null) and form sends following XML text

<dragdot name=A num=N  x=... y=...>
<deldot  name=A num=N+1>

Form is always sent on pressing button "submit".

Drag-and-drog is very convenient to guess of points coordinates.

Work with polygon

click, double, right click

Click, double-click, right-click inside contour of

• count
• broken line, first point of which has coordinates, equal to coordinates of last point
• closed, painted over broken line

<line name=A cmd=... os=... ns=...>
<line name=A cmd=... os=... ns=... user-x=... user-y=...>

If click by mouse right key, then attribute menu is used for tag line (instead of point-menu or section-menu).

drag

Drag-and-drop with pressing mouse inside contour of

• count
• broken line, first point of which has coordinates, equal to coordinates of last point
• closed, painted over broken line

<dragline name=A dx=... dy=...>

Resume

<dot      name= cmd= os= ns=  num= >
<section  name= cmd= os= ns=  num= user-x= user-y= >
<line     name= cmd= os= ns=       user-x= user-y= >
<dragdot  name=      x=  y=   num= >
<dropdot  name=               num= onname= onnum= >
<dragline name=      dx= dy= >
<deldot   name=               num= >

html60author