nix: remove explosions if you have a window size less than four

Turns out strings do not like being resized to -4.

This was discovered while messing with the tests to remove unbuffer and
trying stdbuf instead. Turns out that was not the right approach.

This basically rewrites the handling of this case to be much more
correct, and fixes a bug where with small window sizes where it would
ALSO truncate the attr names in addition to the optional descriptions.

Change-Id: Ifd1beeaffdb47cbb5f4a462b183fcb6c0ff6c524

1 files changed, 28 insertions, 14 deletions

diff --git a/src/nix/flake.cc b/src/nix/flake.cc
index 5ea9e077b..d4eca86b8 100644
--- a/src/nix/flake.cc
+++ b/src/nix/flake.cc
@@ -3,7 +3,7 @@
 #include "common-args.hh"
 #include "shared.hh"
 #include "eval.hh"
-#include "eval-inline.hh"
+#include "eval-inline.hh" // IWYU pragma: keep
 #include "eval-settings.hh"
 #include "flake/flake.hh"
 #include "get-drvs.hh"
@@ -17,8 +17,8 @@
 #include "markdown.hh"
 #include "terminal.hh"
 
+#include <limits>
 #include <nlohmann/json.hpp>
-#include <queue>
 #include <iomanip>
 
 using namespace nix;
@@ -1258,24 +1258,38 @@ struct CmdFlakeShow : FlakeCommand, MixJSON
                             attrPath.size() == 3 && attrPathS[0] == "checks" ? "derivation" :
                             attrPath.size() >= 1 && attrPathS[0] == "hydraJobs" ? "derivation" :
                             "package";
+
+                        std::string output = fmt("%s: %s '%s'", headerPrefix, type, name);
+
                         if (description && !description->empty()) {
                             // Trim the string and only display the first line of the description.
                             auto desc = nix::trim(*description);
                             auto firstLineDesc = desc.substr(0, desc.find('\n'));
-
-                            std::string output = fmt("%s: %s '%s' - '%s'", headerPrefix, type, name, firstLineDesc);
-                            if (output.size() > getWindowSize().second) {
-                                // we resize to 4 less then the window size to account for the "...'" we append to
-                                // the final string, we also include the ' since that is removed when we truncate the string
-                                output.resize(getWindowSize().second - 4);
-                                output.append("...'");
+                            // three separators, two quotes
+                            constexpr auto quotesAndSepsWidth = 3 + 2;
+
+                            int screenWidth = isOutputARealTerminal(StandardOutputStream::Stdout)
+                                ? getWindowSize().second
+                                : std::numeric_limits<int>::max();
+
+                            // FIXME: handle utf8 visible width properly once we get KJ which has utf8 support
+                            //        technically filterANSIEscapes knows how to do this but there is absolutely
+                            //        no clear usage of it that would actually let us do this layout.
+                            int spaceForDescription = screenWidth - output.size() - quotesAndSepsWidth;
+
+                            if (spaceForDescription <= 0) {
+                                // do nothing, it is going to wrap no matter what, and it's better to output *something*
+                            } else {
+                                const char *ellipsis = "";
+                                if (spaceForDescription < firstLineDesc.size()) {
+                                    // subtract one to make space for the ellipsis
+                                    firstLineDesc.resize(spaceForDescription - 1);
+                                    ellipsis = "…";
+                                }
+                                output.append(fmt(" - '%s%s'", firstLineDesc, ellipsis));
                             }
-
-                            logger->cout("%s", output.c_str());
-                        }
-                        else {
-                            logger->cout("%s: %s '%s'", headerPrefix, type, name);
                         }
+                        logger->cout("%s", output);
                     }
                 };