[This file contains a description of the viewer module. See hacking.txt or hacking.html for an overview of the manual.]
As the pager doesn't just follow a static algorithm, but has to react to certain events (keyboard input), the structure is somewhat harder to retrace.
The central pager function is display(), which contains the main loop (keyboard dispatcher) which performs some action(s) on each keypress, as well as the interface to main().
These actions directly performed by the keyboard commands mostly do not have a direct visible effect; instead, they alter the pager state, which becomes visible with the next screen update -- this is performed in each main loop iteration before waiting for the next keypress. Some actions also require temporarily exiting the pager and returning to main(), which is fairly easy and transparent, as the pager state is stored directly in the page structure and the screen contents thus can be easily restored when reentering the pager.
The pager state consists of three basic components: "pager_pos" is the line number (in page) of the first line visible on the screen. "active_link" contains the link number of the currently selected link. (Or -1 if none.) "cursor_x" and "cursor_y" tell where the cursor is on the page. (Not the screen!) The additional "sticky_cursor" flag allows keeping the cursor in its place while scrolling (if possible), instead of moving it to the screen top.
Each of the major parameters is manipulated chiefly by one function: scroll_to() sets "pager_pos", activate_link() sets "active_link", and set_cursor() sets the cursor position (and also manages the "sticky_cursor" status). The variables aren't independent, however -- scroll_to() may need to deactivate the active link if it moved off the screen, and often move the cursor; activate_link may need to scroll if the link is presently not an the screen, and always has to set the cursor position; set_cursor() has to deactivate any previously active link, and also scroll if the desired cursor position is outside the current screen. This interrelation requires the functions calling each other in a quite complicated manner, sometimes even recursively... Maybe we should try to write a generic function for updating all the states at once by a single request.
Note that scroll_to() and activate_link() do not only set the state, but also (re)render the visible page part (using render(), which is described in hacking-layout.*). Probably it would be better to do this in the main loop also. Moreover, activate_link() even modifies the page, which was a very bad idea and surely will be changed in the future.
The viewer basically consists of a keyboard dispatcher that loops until some command was given that requires some action by main(). These include the quit command, link following, history commands, entering the command prompt, or text search.
display() tells main() what action is required, by the return value which is an "enum Pager_ret".
After main() has done it's work, display() will be called again; it will then continue just where it stopped, which is possible because the important pager state information (postion of the currently visible page part, cursor position, currently selected link) are stored in the page structure, which is described under Page List Handling in hacking-page.*. Thus the way over main() is quite seemless; to the user it usually looks as if he was in the pager all the time.
To allow for this, display() needs to do some work to restore the state upon startup.
First, the page is scrolled to the previous position, using scroll_to(). "old_line" is set to PAGE_INVALID (meaning the pager presently displays nothing) before scrolling, so that the screen contents always will be drawn completely anew.
Afterwards, the cursor position is set to the saved one using set_cursor(). Note that the position has to be stored before scroll_to(), as this function might move it out of the screen and adjust the cursor, in case a new cursor position was requested. (By a text search.)
Finally, if there is a selected link, it is (re)activated with activate_link() (described in hacking-links.*). The active link also has to be saved before scroll_to(), as the link position may have changed in the meantime, causing scroll_to() to deactivate the previouly selected link if it is no longer on the screen.
Reactivating the link is omitted when the pager is restarted after a text search command ("search.type" is set) -- in this case the cursor position has higher priority and is not to be overwritten.
Initialization is slightly different if an anchor was activated, which is indicated by "page->active_anchor" being set. In this case, instead of callig scroll_to() directly, activate_anchor() (also described in hacking-links.*) is used. This one also scrolls (but to the anchor position, not the previous pager position), and additionally draws the anchor marks.
After completing these initalizations, the keyboard input loop is entered, which reads one character in each iteration (note that the mvgetch() fuction used for that also updates the curses screen), and then dispatches to the requested command in a big switch.
The cursor position (passed to mvgetch()) is determined from "page->cursor_[xy]", which are page coordinates, and have to be transformed to screen coordinates first.
The scrolling commands are all implemented by simple calls to scroll_to() with different parameters. All relative movement commands can simply add or subtract a constant number of lines to move, as scroll_to() checks for the page boundaries.
This function scrolls the visible page area to the desired position. It moves (or deletes) the screen content, and repaints newly visible areas using render() (see hacking-layout.*).
It is called with the desired new position as argument. This position is described by the line number, of that line of the output page which is displayed in the first screen line.
First the desired new position is checked to be in the valid range.
If it is greater than the page length minus one screenwidth, it is set to that value. (The last page line can't move above the bottom screen line this way.)
If smaller then zero, it is set to zero. (The page top can't move below the screen top.) Note that this will undo the above check, if the output page is smaller than one screenfull -- in this case, the page end is always above the screen end.
Now that we know the real destination position, the difference to the present position (stored in the static "old_line") is calculated.
If the absolute value of that difference is not more than a screenfull, scrolling is used. This is done by the insdelln() curses function, called with the cursor being in the first screen line. (Called with a negative value it deletes lines, causing the screen contents to scroll up; called with a positive argument, it inserts lines, causing the screen contents to scroll down.)
After scrolling, the new lines revealed at the top or bottom of the screen need to be rendered. After scrolling down (negative "scroll_lines"), the first "scroll_lines" of the screen have to be repainted. This is done by calling render() with "scroll_lines" as the height, 0 as the screen position (the first lines of the screen are repainted), and "new_line" as page position. ("new_line" contains the page position of the first screen line.)
After scrolling up, the last screen lines have to be repainted. This is done similar. The hight is "-scroll_lines" ("scroll_lines" is negative when scrolling up), the screen position is the screen end minus "-sroll_lines", and the page position is the page position of the first screen line ("new_line") plus the screen position of the rendered area.
If the difference is too big for scrolling, the whole screen is erased and repainted.
The first time this function is called, "old_line" has the value PAGE_INVALID (defined as the biggest possible positive int value), indicating that the screen contains nothing valid yet. There is no special handling necessary for this case -- "scroll_lines" gets a very big value, and thus the whole screen is always repainted.
Finally, the cursor position is updated, if necessary. If the cursor is no longer on the screen (more exactly, inside the active screen area), it is set to the home position with set_cursor(). If the "sticky_cursor" flag is not set, the cursor is always homed.
Just like the scrolling commands, the cursor movement commands are all implemented by single calls to a specific function (set_cursor()) with the desired new position, which is calculated from the current one.
set_cursor() is responsible for setting the cursor to some desired page position (not screen position!). It also sets the "sticky_cursor" flag, so that the cursor will stay in it's place (again, page position) until reset.
If called with -1 as x and y position, the "sticky_cursor" flag is cleared instead, and the cursor set to the home position -- which is column 0 in the first line inside the active screen area.
If the requested new position is outside the present active screen area, the page first needs to be scrolled. If the position is before the active screen area, we scroll as many lines back as the position is outside; if it is behind the active area, we scroll as many lines forward.
Now we test whether the requested position is on the screen, and truncate if not. The y position is actually tested against the page boundaries (after scrolling, those should be equal in case the requested position is outside); only the x position is tested against the screen boundaries, as there is presently no x scrolling.
Finally, we test whether there is some link at the new cursor position. If there is one, and it wasn't already active, we activate it now. (Using activate_link(), see hacking-links.*.) If there was a link active up to now, but there is no link at the new position, the old one is deactivated. (activate_link() does this automatically, as -1 will be passed as the link number in this situation.)
Only now the cursor is set to the new position, and the sticky status set. (Or reset, if homing.)
As explained under display() above, the pager keeps track of the current active link by the "active_link" variable in the "page" struct. (-1 means no link is active.)
All link selection commands are implemented using a generic link search function to find out which link to activate, and then just activating it with activate_link() (see hacking-links.*).
Finding the link to activate is done with the find_link() function. This function is flexible enough to implement almost all link selection commands with only one invocation.
find_link() looks for the first link or the last link (depending on "match_type") inside a desired range.
The search area is primarily specified by the coordinates "start_x", "start_y" and "end_x", "end_y", enclosing a part of the page inside which the links to be considered have to *start*. (As usual, the start coordinates give the first allowed positin, while the end coordinates give the first invalid position.) Note that regardless of the beginning and end positions being given by coordinate pairs, the range is actually linear: It spanns from the start position to the end of the line, includes (in whole) the following lines till "end_y", and concludes with the beginning of the line "end_y" before the end position.
line zero line one with <zeroth link> in it line two with <firST LINK> IN IT <-- start_y ^start_x LINE THREE WITH <SECOND LINK> IN IT LINE FOUR LINE FIVE WITH <THIRD LInk> in it <-- end_y ^end_x line six with <fourth link> in it
(The second and third links are inside the area, the zeroth, first, and fourth not.)
There is one additional parameter limiting the search area: "end_line" limits the range where links can *end*. (As usual, it gives the first invalid line.) This is necessary as some commands only activate links that are already inside the screen completely.
"end_line" or "end_x" and "end_y" can be specified alternatively, or both -- only links inside both ranges will be considered.
If any of the y paramaters ("start_y", "end_y", "end_line") is given as "-1", it is ignored, i.e. the range extends to the end beginning/end of the page. For the x parameters, the behaviour is more complicated: "-1" actually means that the range starts/ends *before* the beginning of the line; thus, "start_x" being -1 means the whole "start_y" line is inside the range, but "end_x" being -1 means the whole line is *outside* the range, i.e. only links ending *before* "end_y" are accepted. (Just as with "end_line" and other range ends.) This may look confusing at first, but it is consistent: Giving "end_x" means that the beginning of the "end_y" line is inside the range *additionally* to the lines before "end_y", just as "start_x" means the beginning of the "start_x" line is not inside. If this isn't clear, take another look at the figure above, and imagine what happens when "start_x" and/or "start_y" are moved to the beginning of the line.
The implemetation of find_link() is actually fairly small, though not simple.
First, two macros are defined: INSIDE_START() testing whether "link" starts behind the beginning of the range (i.e. whether "start_x" and "start_y" are fulfilled), and INSIDE_END(), which tests the end of the area -- both the end of the link start range ("end_x" and "end_y") and the link end ("end_line").
The first step is finding the candidate link -- when "match_type" is FIND_FIRST, this is the first link starting after the beginning of the range, and with FIND_LAST the last link before the end of the range.
The candidate located using a binary search. (Note that "first" and "last" are really only helper variables for the search, just like "mid". The fact that, depending on "match_type", one of those is actually the result, is only a side effect of the search algorithm; but the respective other one is only garbage. You should *not* confuse them for always being the first and last link inside the area -- only one of the bounds is checked in this search!)
link0 link1 link2 link3 link4 start =>| ^candidate
Now we test whether the candidate is actually a valid link (inside the link number range), i.e. there are *any* links inside the boundary.
Finally, the other boundary has to be tested.
link0 link1 link2 link3 link4 =>| ^ |<= end |------------->| candidate < end => OK
(If there are no links inside the range, still a candidate matching the first bound may have been found.)
link0 link1 link2 link3 link4 =>| |<= ^ |<----| candidate > end => nothing in range
If no link matching the criteria was found, find_link() returns -1.
Cheking whether a link is (or would be) inside the valid screen area for active links, primarily during link selection but also in a few other places, is done using the active_start() and active_end() functions. These simply return the bounds of the valid area (in page coordinates, not screen coordinates!).
Normally, the area starts in the line that shows up at the screen top (i.e. "pager_pos") plus "cfg.link_margin" -- this is one line below the screen top with the default setting. The end is the last line on screen minus the link margin, accordingly. (But note that active_end returns the number of the *first invalid* line, i.e. *after* the end of the valid area, as all end coordinates are specfied this way!)
An exception is the pager being at the top (for active_start()) or bottom (for active_end()) of the page -- in this case the valid area includes the first or last page lines, as these can never move into the normal valid area. (The page top can't ever move below the screen top; same for bottom.)
The "pos" parameter allows specifying some other pager position (relative to the current "page->pager_pos") to check; 0 means using the current position.
Thanks to the powerful find_link() function, the commands are actually very simple. Most commands just consist of one call to find_link() to find the new link, and a call to activate_link() with the new link number to activate it.
The simplest commands are those independant of the current cursor position. ('H', 'L', 'M') These just pass a range of line numbers (aquired with active_start() and active_end()) to find_link().
Many commands however search forward or backward starting from the current cursor position. These just pass the cursor position as the start (for those searching forward) or the end of the search range, respectively. There is one little complication when searching forward: If a link is active already (implying the cursor is at the link start), a search starting with the cursor position will always give the active link again. Thus, "x_start" needs to be incremented by one in this case.
The other end of the range depends on how far the specific command searches: Some (<tab>, 'p') search till the end of the document, while others ('J', 'K', and also '+' and '-') only to the end of the active screen area, optionally with scrolling one line.
The latter also have to test whether a link was actually found in the desired area, and just scroll one line with scroll_to() otherwise.
Some commands ('+', '-', '^', ' and '0') do not directly search from the cursor position, but in a range relative to the cursor line. Those are very similar. Two of them are more complicated however:
'-' is complicated, because it has to find the first link in some range (the previous line having any links), but the beginning of that range isn't fixed -- there might be already links in the previous line, but we might also have to go a few lines back. Before we can search for the desired link, we have to find out the line where to search. This is done with another invocation of find_link(), looking for the last link starting before the current line; as this link is obviously in the previous line having links, we can use this link's starting line as the beginning of our search range.
a line with the <first link> a line with the <second link> and the <third link> <-- start_y=new_link->y_start ~~~~~~~~~~~~~goal ^^^^^^^^^^^^new_link (helper) a line with no links another line with no links a line with the <fourth link> <-- page->cursor_y ^cursor position another line
'0' is complicated, because -- opposite to all other commands -- it has to find a link *ending* in or after some given line. find_link() has no provision for that; extending it in this way just for this single command would be stupid.
Thus, we have to use a trick: We first search for the last link ending *before* the current line (so we can use the existing "end_line" parameter), and then just look for the next link -- which is obviosly the first one ending in or after the current line.
The history commands all work alike:
'b' goes back to the previous page by setting the current position in the history one backwards (decrementing). The page is then (re)loaded by quitting the pager with "RET_HISTORY" -- main() then invokes the necessary actions for loading the page, before restarting display(). (load_page() and init_load() take care for correct loading of pages from history.)
'f' goes forward in the same manner (incrementing "pos").
'B' searches the history backwards, until it finds some page that is followed either by one with the "absolute" flag set or by an internal page (the internal page is treated like a new site). If nothing is found the search stops at the first entry in the history, and that one is taken.
'F' searches forwards, also looking for a page followed by an absoulte URL. The last entry is taken if nothing was found.
'^r' causes the current page to be reloaded, by returning RET_HISTORY, but not changing the position in the page list. Thus it is not really a history command from the user's point of view -- but technically, it is.
'r' searches backwards for a page with "mark" set. It also takes the first page if nothing was found; 'R' does the same forwards.
The marks are set by the 's' command. This one simply sets the "mark" flag of the current "page" structure.
When a link is followed by typing <return>, "RET_LINK" is returned, and the main program follows the current "active_link". This process is described under main() in hacking-links.*.
The 'u' command causes display() to return "RET_LINK_URL"; main() then shows the URL of the currently active link.
'c' returns "RET_URL", and main() shows the current page URL.
'U' returns "RET_ABSOLUTE_URL", causing main() to print the absolute link target URL which is used when the link is actually followed.
When 'q' is typed the pager simply returns with "RET_QUIT", indicating that the main program should quit also.
When ':' is typed the pager quits too, but returns "RET_COMMAND"; this means that the main program should enter command mode.
'/' is very similar: "pager_ret" is set to "RET_SEARCH"; additionally, "search.type" is set to "SEARCH_FORWARD".